# Genspark Help Center
> Guides and answers for every Genspark feature, from Super Agent and AI Slides to membership plans, credits, and account help.
Every Genspark Help Center article as Markdown, in the order of the index at https://www.genspark.ai/llms.txt. Each article starts with its title, then the URL of its HTML page; that page is also served in 18 other languages at https://www.genspark.ai/{locale}/helpcenter/{slug}.
# Super Agent {#super-agent data-doc-group=ai-team}
Source: https://www.genspark.ai/helpcenter/super-agent
## What is Genspark Super Agent? {#what-is-super-agent data-section-type=overview}
Super Agent is Genspark's autonomous AI assistant that thinks, plans, and acts to complete your everyday tasks.
### More than a chatbot {#more-than-a-chatbot}
Instead of just answering questions, Super Agent breaks your request into steps, chooses the right tools, and carries the work through to a finished result — research, content, analysis, design, coding, communication, and more.
### Works with the apps and tools you already use {#works-with-your-apps-and-tools}
Super Agent connects to your everyday apps and data so it can act with real context — pulling from your files, your connected accounts, and your SecondBrain to complete tasks the way you would.
## What's New? {#whats-new data-section-type=changelog}
The new Super Agent is a major step up: stronger, faster, and sharper than ever, it now takes on the kind of work you'd normally hand to a whole team of experts.
### Turn any workflow into a skill {#turning-workflows-into-skills}
Teach Super Agent a workflow once, and it's saved as a skill. Run it anytime, and share it with your whole team.
How to save a Skill: Just tell Super Agent directly — for example, "Save this workflow as a skill" — and the agent handles the rest.
Once saved, your Skill is ready to use in any future conversation. Explore the official skills library or learn more in our Help Center.
Here are two sample projects — click to explore each one:
- [How to create a skill](https://www.genspark.ai/agents?id=7720b50b-43bc-432e-97c5-462c20012185)
- [How to use a skill](https://www.genspark.ai/agents?id=cb2d5181-4d1c-440c-af01-175e11a957ff)
### Built for Complex Work {#built-for-complex-work}
The new Super Agent handles long, multi-step tasks more reliably:
- Sharper on long-running tasks. It reasons through every step without losing the thread. Start it, close your laptop, and come back to finished work.
- Work in parallel. Run several tasks at once in a single project, and let each one build on the others' results.
- Bigger files. Up to 500 MB on the free plan and 1 GB each on paid plans — or pull them straight from Google Drive, AI Drive, and more.
## Classic Super Agent vs. the New Super Agent {#classic-vs-new data-section-type=changelog}
| | Classic Super Agent | New Super Agent |
|---|---|---|
| Environment | No dedicated cloud machine | Dedicated sandbox environment: real browser, own filesystem, and powerful execution |
| Reusability | Re-describe each task | Save any workflow as a reusable skill |
| Long tasks | Handles multi-step tasks primarily in sequence. | Sharper and more reliable, checks own output and build on earlier progress |
## How to Use Super Agent {#how-to-use data-section-type=feature}
Navigate to the Genspark homepage and enter your prompt in the central chat box, or select Super Agent from the "New" menu in the sidebar.
What to ask: Describe any task — research, content, data analysis, design, coding, communication, and more. Super Agent coordinates the right agents and tools automatically.
### Try these examples {#try-these-examples}
**Read more context**
- Give me a weekly recap from this week's meeting notes — wins, decisions, blockers, next steps.
- Use my last product launch deck as a template and create a new one for the November feature release.
**Turn workflows into skills**
- Run a comprehensive industry analysis using Porter's Five Forces framework, then save the workflow as a reusable skill.
- Score a customer's renewal risk from their recent emails, tickets, and usage — save as a skill.
**Research & Analysis**
- Explain the differences between Web3 and traditional web applications in simple terms
- Research electric vehicle market trends and create a 10-slide presentation with key insights
**Data Analysis & Visualization**
- Analyze the sales data from my uploaded file and build a category analysis with charts and an average-order-value table
**Coding & Development**
- Create a product landing page for my smart water bottle based on the uploaded product brief
**Image & Video Creation**
- Create four different style birthday posters for my friend Vivian's 30th celebration
**Communication & Coordination**
- Draft a professional email to schedule a marketing meeting about the New Year campaign (Connect Gmail or Outlook first to enable sending)
**Download to AI Drive**
- Download 10 influential 2024 research papers on solar energy and save them to my AI Drive
## Files & Sessions {#files-and-sessions data-section-type=feature}
In New Super Agent mode, files the agent creates or works with are kept in the workspace for the duration of your session, so it can build on them across multiple messages.
Anything the agent delivers (the file cards you see in chat) is saved with your project and stays downloadable.
💡 If there's something you want to keep, ask the agent to deliver it rather than leaving it in the working environment.
Share files individually: you can now share each delivered file on its own, separately from the conversation — so you can send a single file to anyone without sharing the whole project.
## FAQ {#faq data-section-type=faq}
### Does New Super Agent use more credits? {#new-super-agent-credit-usage}
Not because of the mode itself. A New Super Agent task can use more credits overall because it does more — multi-step work, live browsing, running code, and more tool calls.
To keep credit use low:
- Choose a lower model tier when you don't need maximum power.
- Nail the prompt in AI Chat first. Use AI Chat to shape and refine exactly what you want, then hand the finished prompt to Super Agent — so it executes cleanly instead of spending credits on back-and-forth clarifying steps.
# AI Slides {#ai-slides data-doc-group=office-suite}
Source: https://www.genspark.ai/helpcenter/ai-slides
## What is AI Slides? {#what-is-ai-slides data-section-type=overview}
**AI Slides** is Genspark's intelligent presentation agent. Instead of just generating a deck and handing it to you, it builds the deck *around you* — following your needs, your content, and your brand.
- **Built around you** — Already have a template or brand style? It follows them. Want something new? It designs from scratch.
- **Conversational by design** — It understands your goal and talks it through with you first, then structures and builds the deck.
- **A full presentation workspace** — Your deck builds live on a canvas beside the chat. Point at anything to change it, queue up batch edits, verify the facts, add speaker notes, and present with a speaker console.
- **Expert Skills, not just templates** — Apply 100+ built-in expert Skills, or save your own style as a reusable Skill.
- **A real built-in computer** — It can actually run code, so data analysis, calculations, and charts are computed accurately rather than guessed.
- **Multimedia and document conversion** — Add AI-generated or web images, and import existing files (PDF, Word, Excel, PPT) to build from.
**Learn more about AI Slides:** https://www.genspark.ai/blog/genspark-ai-slides
## Quick Start {#quick-start data-section-type=overview}
Get started in three simple steps:
1. **Go to AI Slides**
Click "AI Slides" at the bottom of the Genspark homepage, or select it from the "New" menu in the sidebar.
2. **Describe what you want to create**
Enter your topic, goal, or key message. You can also upload documents for AI to analyze and extract key insights.
Sample prompts:
```trythisprompt
Create a 10-slide presentation on renewable energy trends in 2024
```
```trythisprompt
Build a pitch deck based on my uploaded business plan
```
```trythisprompt
Make a sales presentation for our new SaaS product
```
3. **Get your presentation**
AI Slides automatically researches, structures, designs, and generates a complete, visually compelling deck, building it page by page in the Slides workspace right beside the chat. Review, refine, and export as PDF, PPTX, or Google Slides.
## Choose Your Mode {#choose-your-mode data-section-type=feature}
Before you write your prompt, take five seconds to set the modes at the top of AI Slides. There are three choices, and they make a big difference to your result.
### Professional vs Creative {#creative-mode}
This sets the overall style direction:
- **Professional** — Clean, structured, business-oriented. Best for sales proposals, internal reports, and consulting decks.
- **Creative** — More visual impact, with richer colors and imagery. Each slide is designed as a complete visual, like a poster. Best for brand campaigns, product launches, and marketing materials.
Quick rule: presenting to clients or executives → **Professional**; presenting to consumers → **Creative**.
A few things to know about **Creative Mode**:
- It supports multiple aspect ratios (16:9, 4:3, 1:1, 3:4, 9:16, 21:9), making it suitable for posters, social media carousels, knowledge cards, and more.
- You can choose the image generation model (e.g. GPT Image 2 or Nano Banana Pro) and paste reference images to guide the style.
- **Draw to edit** is on by default — mark anywhere on a slide with the pen, marker, or box tool, then describe your change. AI regenerates just that part while keeping the rest intact.
> **Note:** Once a project is created, its mode cannot be changed. To switch modes, start a new project.
### Guide Mode {#guide-mode}
By default, you type one prompt and get instant slides — fast and effortless. **Guide Mode** takes a different approach: AI acts as your presentation consultant, walking you through a structured consultation before building a single slide.
The difference is the role it plays. In the default mode, AI acts like an executor — once it understands your request, it builds the deck as fast as possible. In Guide Mode, it acts like a co-creator — a professional mentor that keeps discussing with you, sparks ideas, and confirms the outline with you before generating the final deck.
**How to choose:**
- Just want a quick result → leave Guide Mode off.
- New to AI Slides, building something complex or formal, or want control over every detail → turn Guide Mode on.
### Standard vs Ultra {#standard-vs-ultra}
This sets the model quality:
- **Standard** — Fast and economical, good for everyday decks. Uses 0.5× credits using [Genspark Gen-1 model](https://www.genspark.ai/helpcenter/gen-one-model).
- **Ultra** — The best choice for high-quality decks, with stronger reasoning and output. Uses 1.0× credits.
Quick rule: choose Standard when you want a fast deck, Ultra when you want the best quality.
> **Enterprise model policies:** if your organization restricts AI model vendors ([Team Admin → Connectors](/payment/team_admin?view=connectors) → **"Restrict AI Model Vendors"**), AI Slides automatically picks every model it uses — including the Standard and Ultra tiers — from your approved vendors. Members don't need to change anything.
## Writing Effective Prompts {#writing-prompts data-section-type=feature}
The key to a great deck is writing your prompt from the outcome. A vague prompt like "make a deck" leaves AI guessing who it is for, how long it should be, and what to cover.
### Start with three essentials {#prompt-essentials}
Open your prompt with three things:
1. **Audience** — who is it for? (Clients? Team? Executives?)
2. **Length** — roughly how many slides?
3. **Structure** — how should the content be organized?
Instead of "make a marketing deck", write something like:
```trythisprompt
Build a Gen Z marketing strategy deck, 5 slides, conclusion first, covering market analysis, target audience profile, and marketing strategy
```
The more specific you are, the sharper the output.
> **Not sure how to phrase it?** Just turn on Guide Mode and leave the rest to AI — it will walk you through and clarify what you need.
### Add more detail when it helps {#prompt-advanced}
- **Data source** — "use 2024 data", "cite authoritative sources"
- **Tone** — "professional business", "young and energetic", "minimalist"
- **Reference style** — paste a screenshot of a deck you like
- **Your own files** — drag in a PDF, Word, Excel, or PPT and build from it
### Prompt templates for common scenarios {#prompt-templates}
- **Sales proposal** — "Build a sales proposal from [your company] to [client company], [N] slides, conclusion first, emphasizing the product value for the client."
- **Internal report** — "Build a [topic] quarterly review report, [N] slides, structured as current state, achievements, issues, and next-quarter plan."
- **Training material** — "Build a [topic] training deck, [N] slides, for beginners, with a hands-on case per chapter."
## Slide Skills {#slide-skills data-section-type=feature}
Slide Skills are AI Slides' most powerful way to get expert-quality decks and keep your style consistent. Skills replace the older idea of templates — and go far beyond them.
### What is a Skill? {#what-is-a-skill}
A traditional template only reuses a visual look. A **Skill** packages **expert thinking + design** together. It tells AI not only how the deck should look, but also how to think, how to organize the content, and how to build the argument. Think of it as the way an expert works, packaged for reuse.
### Use built-in Skills {#using-skills}
Genspark ships with **100+ expert Skills**, covering a wide range of work scenarios — Corporate Strategy, B2B Sales, Professional Training, Academic Research, Marketing, Data & KPI, Fundraising, Consulting, Product Management, and more.
1. On the AI Slides home screen, browse the **Skills** library below the input box. Switch between **Featured** and category tabs to explore.
2. Click the Skill you want to apply.
3. Enter your prompt — AI generates the deck following that Skill's expert approach.
With a Skill applied, the structure, the reasoning, and the visual style all follow how an expert would actually do it: a sales proposal uses a sales director's framing, a strategy analysis uses a consultant's frameworks.
### Save your own Skill {#save-your-skill}
Beyond the built-in library, you can save your own style as a reusable Skill — especially useful for reusing a brand template across your team. Your saved Skills appear under the **My Skills** tab.
The most common case is that **you already have your own PowerPoint or PDF template.** Open the **New Skill** dialog from the Skills library and choose how to start:
- **Save my presentation as a template** — upload a PowerPoint or PDF to turn your existing deck into a reusable Skill. This is the path most people use.
- **I already have a skill package** — import a `.zip` Skill that someone shared with you.
- **I'm starting from scratch** — build a custom Skill from the ground up.
You can also save a deck you are currently working on: click the **⋮** menu at the top of the workspace → **"Save as Skill"**.
**For client work:** you can do the same with a client's brand — look up their colors and style, save it as a Skill, and apply it whenever you build a deck for them. It is a small touch that makes clients feel understood.
> **Reuse a project on the fly:** You can also start a new project, paste the link of a project whose style you like, and tell AI to use it as the basis for your new deck.
### Share a Skill {#share-skills}
Every Skill you own has a **Share** panel — share it like a doc, with one link, instead of exporting a `.zip` and sending files around:
1. Open one of your Skills and click **Share**.
2. Invite specific people by email, or set General access to **"Anyone with the link"** and copy the link.
Recipients can preview the Skill and import their own copy. You stay in control — remove a person or disable the link at any time; copies people already imported remain theirs.
### Team Skills: publish to your whole organization {#team-skills}
Person-to-person sharing spreads a Skill one colleague at a time. **Team Skills** publish it to everyone at once, so your whole team builds decks one consistent way — the company proposal framework, a department's brand style, a report format that took years to refine. A new employee can build a polished, on-brand deck on day one. Available on both the Team and Enterprise plans.
1. **Members** — open your Skill, click **Share**, and set General access to **"Organization"**. This submits the Skill for admin review.
2. **Admins** — review submissions in [Team Admin → Skills](/payment/team_admin?view=skills&skilltab=slides) and approve or reject (with a reason). Email and in-app notifications keep the loop moving on both sides.
3. Once approved, the Skill appears in the **"Team Skills"** tab of the Skills gallery for every member and activates automatically — no importing needed. When the publisher updates it, everyone gets the latest version.
## Charts & Data Visualization {#charts data-section-type=feature}
Presentations often stall on diagrams — process flows look uneven, org charts get tangled, and data charts mean a separate trip to Excel. AI Slides handles all of this internally, and because it actually runs code, the numbers behind your charts are computed accurately.
### Charts you can generate directly {#chart-types}
- **Business process flows** — order-to-delivery, user journeys
- **Org charts** — clean hierarchies
- **2x2 matrix analysis** — competitor positioning, prioritization
- **Funnel charts** — sales and conversion funnels
- **Gantt charts and timelines** — project schedules, milestone planning
- **Data charts** — bar, line, pie, radar, and more
These are just the most common types — basically any chart you can think of, AI Slides can produce.
### How to generate them {#chart-how}
Specify what you want directly in your prompt:
```trythisprompt
Build a 10-slide product launch plan. One slide needs a 2x2 competitor analysis (by price and functionality), and another needs a user growth funnel.
```
You can also hand AI a data file and ask it to visualize it:
```trythisprompt
Use this sales spreadsheet to create a monthly trend line chart, a product-line comparison bar chart, and a regional distribution heatmap
```
> **Not sure which chart fits?** Just ask — "What is the best chart for this content?" — and let AI recommend one. It often picks the most fitting type, sometimes one you would not have thought of.
## Refining Your Slides {#refining data-section-type=feature}
Your deck lives in the **Slides workspace**: an outline rail on the left shows every slide at a glance, the current slide sits center stage, and the chat stays right beside it — so you can talk to AI and watch every change land in real time. Above the slide, a toolbar switches between three ways of working: **Select**, **Draw**, and **Edit**.
### Point at what you want changed {#select-edit}
In **Select** mode, click any element on a slide — a title, a paragraph, an image, a chart. The element attaches to the chat box, labeled by type; describe what you want, and AI changes exactly that spot while keeping the rest of the design intact.
### Mark up the whole deck, send once {#draw-edit}
Review your deck the way you would go through a printed draft. Switch to **Draw**, then circle things with the **Pen**, highlight with the **Marker**, or drag a **Box** around a region — and jot a short note on each mark. Annotations stay put as you flip through pages, collecting into the **Edit queue** at the chat box. When you have marked up everything, click **"Send N edits"** once: AI works through the whole batch while you grab a coffee or join your next meeting, and you come back to review the results in one pass.
Changed your mind about a mark? Remove it from the queue before sending, or click **"Discard"** to clear the current page or all pages.
### AI Edit {#ai-edit}
The toolbar also offers one-click AI refinements. The available options depend on your mode:
- **Professional Mode** — **Fix Layout** (detect and fix overlap, overflow, and misalignment) and **Polish Content** (improve logic and layout while matching the current style).
- **Creative Mode** — **Enrich Content**, **Simplify Content**, and **Redesign Layout**.
### The built-in editor — full manual control {#advanced-edit}
Sometimes you don't want to ask AI — you just want to grab that title and move it yourself. Switch to **Edit** in the toolbar and the slide becomes directly editable, right on the canvas: no page change, the chat and AI stay by your side, and manual edits do not consume credits.
- **Work with elements like a design tool** — click to select, drag to move, use the handles to resize and rotate. Marquee-drag to select several elements, then align, distribute, or group them, and reorder layers (bring forward / send back). Right-click any element for cut, copy, paste, and duplicate.
- **Format text in place** — double-click into any text and type. The floating toolbar covers font, size, color, bold/italic/underline/strikethrough, alignment, line height, and letter spacing.
- **Fine-tune images** — replace an image, crop it freely or to a shape or ratio, round its corners — or open **Change Focus** and drag a dot on a miniature of the original photo to choose exactly which part stays in frame.
- **Manage pages from the rail** — drag thumbnails to reorder; right-click a page to add, duplicate, copy, paste, or delete slides.
- **Edit speaker notes** — type directly in the **Notes** panel while you edit.
- **Undo anything** — undo any action in a single step (and redo just as easily), and the **History** panel keeps full version snapshots; manual edits and AI edits share the same history. Changes save automatically, and if the deck changes while you are editing — say, AI finishes an edit — you will be shown the newer version and choose how to proceed, with no silent overwrites.
*(The built-in editor is only available in Professional Mode)*
### Verify Content {#fact-check}
Before an important deck goes out, have AI verify every figure and claim in it — checked against the web or against the files you uploaded:
1. Click **"Verify content"** in the toolbar and choose **"Verify this page"** or **"Verify entire deck"**.
2. AI cross-checks facts, figures, sources, and calculations, corrects what is wrong, and leaves a verification mark on each checked slide with reference links.
3. Review the marks to see what was checked and how. When you export to PPTX, the verification trail is written into the slide notes, so whoever receives the file sees the evidence behind every number.
This step is especially valuable for sales proposals and reports, where solid data matters most. A reliable flow is: **simple prompt → fast generation → verify → sharp result**.
## Present, Export & Share {#present-export-share data-section-type=feature}
### Speaker notes {#speaker-notes}
Once the deck is ready, shape your delivery. Open the **Notes** panel from the bottom toolbar to read and edit per-slide speaker notes — or have AI write them: click **"Add notes"** (for this slide, or for all slides), and AI first asks a few questions about the occasion, the audience, and the tone, then writes notes tailored to your answers.
Notes are included when you export to PPTX, so anyone you hand the file to can present it smoothly too.
### Present online {#view-online}
Present your slides directly in the browser, with no download required:
1. Click **"Present"** at the top of the workspace.
2. Choose **"From beginning"** or **"From current slide"**.
3. While presenting, use **"Draw on page"** to circle and underline things live for your audience.
### Presenter view {#presenter-view}
For the full on-stage setup, click **"Present"** → **"Presenter view"**. A speaker console opens in its own window with the current slide, the next slide, your notes, and a timer — while the audience window shows nothing but clean slides. Move the audience window to the projector or the shared screen, make it fullscreen, and drive the whole show from your console; your cursor stays synced across both displays.
### Export {#export}
Download your presentation as PDF, PPTX, or Google Slides *(available for Plus, Pro, Team, and Enterprise users)*:
1. Open your presentation, then click **"Export"** in the deck header's right-hand action row. Export appears only when a deck is open and editable — an embedded preview or an anonymous share viewer has no Export button.
2. In the **"Export Slides"** dialog, choose a **Format** - PDF, PPTX, or Google Slides - and set a **page range**: All pages, or Custom range with From/To. Export stays disabled on an invalid range.
A quick decision rule:
- Still need to edit it → export **PPTX** to PowerPoint
- Sending it straight to someone → export **PDF**
- Collaborating with a team → export **Google Slides**
### Make a Copy {#make-a-copy}
Create a full, independent copy of your presentation as a new project — useful when adapting slides for a different client, translating into another language, or creating a variation without affecting the original. Click the **⋮** menu at the top of the workspace → **"Make a Copy"**, or simply tell AI **"make a copy"** in the conversation.
### Share {#share}
Share your presentation through Genspark's project sharing:
1. Open the project share options.
2. Invite collaborators by email, or change General Access to **"Anyone with the link"**.
3. If you choose "Anyone with the link", pick your sharing mode:
- **Presentation share** — final slides only; the creation process is not included.
- **Project share** — includes the full project and creation process.
### Editing in Canva or Figma {#canva-figma}
Want more design control? Export your slides and continue editing in Canva or Figma for advanced customization.
**Learn more:** [External Editing Guide](https://www.genspark.ai/blog/genspark-ai-slides-edit)
## Tips & Saving Credits {#tips data-section-type=feature}
### Plan for clean, balanced layouts {#clean-layouts}
Want professional-looking slides? Plan your content structure before generating. Start with an outline and keep the information volume consistent per slide. If a slide looks crowded, split it across multiple slides — you can always pause generation and adjust as you go. Try:
```trythisprompt
Create a comprehensive outline first, determine the content for each slide, and keep the information volume consistent per slide. Wait for my confirmation before starting production.
```
### Keep your style consistent {#style-consistent}
Two approaches that work well:
**1. Sequential creation** — tell AI to generate slides one at a time:
```trythisprompt
Please create slides one by one, do not process in parallel
```
**2. Style approval workflow** — approve the first slide, then have AI match it:
```trythisprompt
Please create the first slide and wait for my approval of the style. Then create all remaining slides to match that exact style — same fonts, colors, spacing, and layout structure.
```
For long-term consistency, save your approved style as a [Skill](#slide-skills).
### Save credits {#saving-credits}
AI generation is what consumes credits — manual edits in **Advanced Edit** are free. A few ways to make the most of your credits:
1. **Think first in a free AI chat.** Building a deck spends credits on two things: thinking through the content, and producing the slides. You can do the "thinking" part for free — talk through your content and per-slide plan in a free AI chat (like Genspark's AI Chat), then hand the draft to AI Slides to produce. You only pay for production.
2. **Create with Ultra, reuse with Standard.** Building a high-quality deck from scratch is worth the stronger model, so use **Ultra** when creating a reusable Skill. Reusing a finished Skill is far easier, so switch to the cheaper **Standard** for everyday decks built from it.
3. **Stop early.** If the output isn't what you want, stop generation immediately instead of letting unused slides finish.
4. **Use Advanced Edit for small changes.** Reserve AI generation for major revisions; use the free Advanced Edit for text, formatting, color, and layout tweaks.
## FAQ {#faq data-section-type=faq}
### Why won't my presentation export, download, or preview? {#export-troubleshooting}
Try these steps in order:
- **Refresh and retry.** Reload the page and reopen the presentation, then attempt the export again.
- **Check your browser.** Disable pop-up blockers for genspark.ai and make sure downloads aren't being blocked. Try Chrome or Edge if the issue persists.
- **Try a different format.** If PPTX won't export, try PDF (or vice versa) to narrow down where the issue is.
If it still won't export, contact support and share your project with **support@genspark.ai** so we can look at it directly (see below).
### Do exports cost credits? {#export-credits}
No — exporting does **not** cost credits. Neither does manual editing (Advanced Edit) or saving a presentation. Credits are only consumed by AI generation.
### I made a deck on a Free plan, how do I get the file {#free-plan}
Exporting any format, requires a paid plan.
### How do I share my project link with support? {#share-project-link}
On Free you can present in the browser and share a live link, but downloading needs an upgrade. Your deck is saved — nothing is lost — and it becomes exportable the moment you upgrade.
Support can only investigate a presentation if the project link is set to public.
- Open the agent or task containing your presentation.
- Click **Share** (top right of the project).
- Invite **support@genspark.ai** as a Viewer
- Copy the link and paste it into your support email.
> ⚠️ Without public access, we can't open the project — this is the most common reason investigations get delayed.
# AI Sheets {#ai-sheets data-doc-group=office-suite}
Source: https://www.genspark.ai/helpcenter/ai-sheets
## What is AI Sheets? {#what-is-ai-sheets data-section-type=overview}
AI Sheets is Genspark's autonomous AI spreadsheet agent that turns data analysis into natural conversation. Simply describe what you need — AI handles data collection, analysis, and visualization automatically.
- **Multi-Source Data** — Web scraping, file uploads, or database connections in one prompt
- **Autonomous Analysis** — Clean messy records, compare performance, forecast trends, discover patterns, calculate metrics, etc.
- **Native Excel Compatible** — Full support for .xlsx format with seamless import/export and generation of standard Excel formulas
- **Smart Visualization & Styling** — Transform raw spreadsheets into polished, presentation-ready tables with formatting and design, generate professional charts based on your needs
- **Actionable Insights** — Extract key findings and strategic recommendations for decision-making
## Quick Start {#quick-start data-section-type=overview}
Get started in three simple steps:
1. **Go to AI Sheets:**
Click "AI Sheets" at the bottom of the Genspark homepage, or select it from the "+" (New) button in the sidebar.
2. **Describe your needs:**
Tell AI what you want to analyze. Sample prompts:
```trythisprompt
Compile NVIDIA's financial statements from FY2021 to FY2025.
```
```trythisprompt
Analyze sales trends from the uploaded Excel and visualize performance by region.
```
3. **Get results:**
Based on your needs, AI automatically selects and executes the right analysis steps — cleaning messy data, writing SQL queries for databases, running Python for statistical analysis, generating Excel formulas, or creating visualizations. All code and logic are transparent. Edit directly in the sheet, refine with follow-up prompts, or export as .xlsx for further use.
## Feature & How-To Guides {#feature-guide data-section-type=feature}
### Data Collection {#data-collection}
AI Sheets automatically collects structured data from authoritative sources:
* **Professional Databases:** Genspark integrates specialized databases across different domains:
* **Financial Data** — SEC (public company filings, 10-K/10-Q, etc.), Yahoo Finance (stock prices, etc.)
* **Academic Data** — Google Scholar (papers, citations, research trends, etc.)
* **Product Data** — Amazon (product information, pricing, reviews, etc.)
* **General Web Data:** For non-database data, AI retrieves high-quality information from public web sources based on your needs, and cross-checks critical data against multiple sources. Verification still matters — always check critical values against the source cells and citations before you rely on them.
AI Sheets supports data extraction and analysis from multiple sources:
* **Upload Files via Chat:** Upload local files directly (Excel, PDF, Word, images, etc.), and AI will automatically extract and analyze the data.
* **Connect Data Sources via MCP:** On the Genspark homepage or within AI Sheets, click the "Tools" button in the chat box, open the MCP page, select the tool you want to connect, and follow the instructions to complete the setup. Once connected, you can use prompts to have AI directly query the data source and perform analysis. Currently supported:
* **Google Drive** — Access cloud files, extract documents and spreadsheet data
* **Supabase** — Connect to your database, execute SQL queries and data analysis
* **Custom MCP** — Other MCP tools that meet AI Sheets query requirements, which can connect to enterprise databases
For the full connector list and how connections work, see [Connectors and integrations](https://www.genspark.ai/helpcenter/connectors-and-integrations).
### Sheet Operations {#sheet-operations}
AI Sheets features a built-in web-based spreadsheet editor that supports both manual editing and AI-assisted editing:
* **Manual Editing:** You can operate the spreadsheet just like Excel:
* Copy, paste, cut, delete data
* Use common Excel formulas for data processing and analysis
* Create and adjust charts, including: Area charts, Bar charts, Column charts, Line charts, Pie charts, Doughnut charts, Scatter charts, Funnel charts, Radar charts, Treemap charts, Waterfall charts
* Format settings and style adjustments
* Add and delete worksheets
* Click **Fullscreen** in the upper right corner to enter full-screen editing mode for more workspace.
* **AI Editing:** You can also direct AI to perform spreadsheet operations using natural language:
* **Generate Formulas** — Describe calculation logic in natural language, and AI automatically generates the corresponding Excel formula:
```trythisprompt
Show what percentage each product contributes to overall revenue
```
```trythisprompt
Calculate the compound annual growth rate (CAGR) from 2020 to 2024
```
```trythisprompt
Add a column that shows "High" if sales > 10000, "Medium" if between 5000–10000, otherwise "Low"
```
* **Generate Charts** — Describe visualization needs, and AI auto-selects the appropriate chart type and generates it:
```trythisprompt
Visualize the market share comparison between competitors using a pie chart
```
```trythisprompt
Generate a bar chart comparing sales performance across different regions
```
```trythisprompt
Show customer satisfaction performance by department using a radar chart
```
* **Focused Area Editing (Send to Chat)** — After selecting a cell range, click the "Send to Chat" button to send the selected area to the chat box. You can direct AI to perform content corrections, in-depth analysis, or chart generation for that specific area:
```trythisprompt
Analyze the selected data and identify the top 3 anomalies or outliers
```
```trythisprompt
Highlight duplicate values in the selected customer ID column
```
```trythisprompt
Standardize all phone numbers in the selected range to (XXX) XXX-XXXX format
```
### Analysis & Insight {#analysis-insight}
You can ask questions about spreadsheet content, and AI will provide answers with cell references. You can also have AI deeply explore the data and generate insights, with the agent creating beautiful visualization dashboards using HTML tools.
* **Share:** Click "Share" in the top right corner, adjust permission settings, and share your project link for viewing or collaborative editing.
* **Export:** Click **Export** in the upper left corner of the editor, then choose **Export as XLSX**. Your formatting, formulas, and calculations are preserved, and the file opens in Microsoft Excel and other spreadsheet apps.
## Credits and Free Quota {#credits-and-free-quota data-section-type=feature}
AI Sheets works through agent tasks. A few things are worth knowing before you spend:
* **A run that partly succeeds still costs credits.** Attach the files the task needs and give clear, well-scoped instructions before you start. If a task keeps failing on the same step, stop it and fix the input rather than retrying.
* **Free Quota covers some runs.** If your plan includes Free Quota — the weekly allowance that covers eligible tasks at no credit cost — a covered AI Sheets task is paid from the Free Quota pool instead of your credits. See How Free Quota Works for what's covered.
* **Billing questions.** For a technical billing issue, contact support with a link to the task, what went wrong, and your approximate balance before and after. For how credits are counted and charged, see the [Credits Guide](https://www.genspark.ai/helpcenter/credits-guide).
## Pro Tips {#pro-tips data-section-type=feature}
### Break Down Complex Tasks into Steps {#break-down-tasks}
* For multi-step analysis tasks, have AI complete data collection and cleaning first, then proceed to in-depth analysis
* Split complex tasks into multiple instructions, verify results after each step before telling AI to proceed to the next step to avoid errors that could impact the final results
**Sample: Financial Analysis on NVIDIA**
* Collect NVIDIA's revenue, net income, and operating margin data from FY2020 to FY2024.
* Calculate year-over-year growth rates for revenue and net income, and visualize the trends.
* Analyze what drove the revenue growth in FY2024, compare with industry benchmarks, and provide strategic insights.
**Sample: ROI Analysis of Marketing Campaign**
* Use my marketing campaign data and calculate total spend, conversions, and revenue generated by each campaign.
* Based on the results, calculate ROI and cost per conversion for each campaign, then rank them by performance.
* For the top 3 campaigns, analyze which channels and audience segments performed best, and recommend budget allocation for next quarter.
### Multi-Sheet Collaborative Analysis {#multi-sheet-analysis}
When working with complex data that spans multiple dimensions, leverage AI Sheets' multi-tab capabilities to understand existing workbook structures and perform cross-sheet relational analysis automatically.
* Upload existing multi-tab spreadsheets and ask AI to explain complex formulas and model structures
* Create multiple sheets within a single project for organized data management
* Use natural language to direct cross-sheet analysis:
* Link sales data from Sheet1 with inventory data from Sheet2 to generate a restocking recommendation table
### Reduce Conversation Rounds per Project {#reduce-conversation-rounds}
* When there are too many conversation rounds, AI may lose performance due to context compression
* Whenever the context becomes too long, create a new Sheets agent project
## AI Sheets vs GenOffice vs Genspark for Microsoft Office {#ai-sheets-vs-genoffice-vs-genspark-for-microsoft-office data-section-type=feature}
Three different products share the spreadsheet name:
* **AI Sheets** (this page) — Genspark's autonomous AI spreadsheet agent
* **GenOffice** — a separate, standalone open-source office suite (Docs, Sheets, Slides, PDF) with AI built in
* **Genspark for Microsoft Office** — a set of three add-ins for Word, Excel, and PowerPoint that open an AI task pane inside the app
If you're working inside Microsoft Excel and want AI in the task pane, see [Genspark for Microsoft Office](https://www.genspark.ai/helpcenter/microsoft-office-plugin).
## Troubleshooting {#troubleshooting data-section-type=troubleshooting}
### 1. It said the task was done, but the sheet didn't change {#it-said-the-task-was-done-but-the-sheet-didnt-change}
* Check you're looking at the right place. If the project has more than one sheet (tab), confirm the change landed on the one you're viewing.
* Name the exact cell instead of describing the problem. "Cell B12 shows 0; it should be 1,240" works better than "fix the totals".
* Give one instruction at a time, and check the result before sending the next one.
* If an earlier step went wrong, restore the earlier version — see [3. A sheet went blank, or I lost a step](#a-sheet-went-blank-or-i-lost-a-step).
### 2. The export didn't download {#the-export-didnt-download}
* On desktop, use **Export** in the upper left corner of the editor, then choose **Export as XLSX**. There is no separate "Download" item.
* "Completed" means the task finished — it doesn't guarantee your browser saved the file. Check your Downloads folder, and if nothing appears, allow downloads for genspark.ai in your browser.
* If the file still doesn't download, open the project on a desktop browser and try **Export as XLSX** again.
### 3. A sheet went blank, or I lost a step {#a-sheet-went-blank-or-i-lost-a-step}
* On the result card in the chat, click **Restore** (the tooltip reads "Restore to this version"). It reloads that earlier version into the live workbook and saves it again.
* **Restore** takes the workbook back to the state before that step. It isn't a browsable version history, and you can't name or keep separate snapshots of a workbook.
* The button is greyed out with "Cannot restore while AI is responding" — wait for the current task to finish, then restore.
* To protect a good state, use **Export as XLSX** on desktop before you try something risky.
### 4. A chart or format didn't update {#a-chart-or-format-didnt-update}
* Name the object and the sheet: "change the pie chart on the Summary sheet to show Q3".
* If it still looks unchanged, restore the earlier version and send the instruction once more.
* Charts are supported — including area, bar, column, line, pie, doughnut, scatter, funnel, radar, treemap, and waterfall. An unchanged chart usually means the instruction didn't reach the right chart.
### 5. The output language changed {#the-output-language-changed}
* AI Sheets has no output-language setting, and changing your Display Language doesn't change its output — the agent replies in the language of your last message.
* To change it, write your next message in the language you want, or simply ask it to switch.
### 6. A deleted project can't be recovered {#a-deleted-project-cant-be-recovered}
Deleting an AI Sheets project is permanent. There is no undo, and no way to restore a deleted project from the project list. Before you delete, use **Export as XLSX** to save anything you might need.
## FAQ {#faq data-section-type=faq}
### Can AI Sheets be used on mobile devices? What features are supported on mobile? {#mobile-support}
AI Sheets currently only supports viewing results on mobile devices. We recommend using a desktop computer for full functionality.
### Can AI Sheets use specific statistical methods (such as regression analysis, hypothesis testing) for analysis? {#statistical-methods}
AI Sheets supports most data analysis that can be accomplished through Excel formulas, including statistical analysis.
### What is the data limit that AI Sheets can support? {#data-limit}
AI Sheets doesn't enforce a fixed cell limit — there's no set maximum number of cells. Practical performance depends on your computer's hardware and available memory: on desktop, AI Sheets performs spreadsheet operations in your browser, so keep enough machine resources free to avoid lag and memory overload. If a large workbook starts to slow down or use too much memory, split it into smaller sheets or start a new project. When the agent reads a very large range, it may process it in pieces — a single range read is capped at 3,000 cells.
### How accurate is the data AI Sheets collects? {#how-accurate-is-the-data-ai-sheets-collects}
AI Sheets cross-checks critical data against multiple sources, but automated collection isn't perfect. Before you rely on a value, verify it against the source cells and citations in the sheet — especially for high-stakes decisions.
### What should I do if AI Sheets makes an error? {#ai-sheets-error}
Tell AI the exact error and how it should be corrected, export and back up your results as you go, and break complex tasks into a series of simpler tasks. For tasks that need high accuracy, review all changes before you finalize the output.
If you can't resolve the issue, or you have a specific feature request, contact support. When support asks to see the project, share it with support@genspark.ai (in the Share dialog, invite support@genspark.ai by email as a Viewer) rather than making it public.
### What security considerations should I be aware of when using AI Sheets? {#security-considerations}
We prioritize data security. Your data is private and is never used for other purposes.
User Best Practices:
* **Source Validation:** Be cautious with files from untrusted sources, as hidden content could mislead the AI.
* **Access Control:** Please ensure project sharing permissions are set correctly to prevent unauthorized access.
### What Excel features are incompatible with AI Sheets? {#excel-incompatible}
The following are currently not supported:
* Pivot tables of any kind, including Power Pivot, aren't supported.
* **Macros & VBA execution:** You cannot execute VBA scripts natively within AI Sheets. However, AI can generate VBA code for you to copy and run in desktop Excel.
* **Excel native data connections:** Legacy connections (ODBC, OLEDB, external links) are not supported.
* **Complex or legacy charts:** Such as 3D charts or specific specialized diagrams. Standard charts are supported.
* User-defined functions (UDFs).
* Power Query.
# AI Docs {#ai-docs data-doc-group=office-suite}
Source: https://www.genspark.ai/helpcenter/ai-docs
## What is AI Docs? {#what-is-ai-docs data-section-type=overview}
**AI Docs** is Genspark's intelligent document creation agent that transforms how you create, edit, and optimize documents.
* **Smart AI assistance** Latest LLMs support document generation and modification through interactive prompts
* **Four output formats** Rich Text, Markdown, Web Doc, and Word meet different professional needs
* **Automated design intelligence** Automatically applies professional design principles, typography hierarchies, and logical structure
* **Visual integration** Upload images for AI to analyze aesthetic elements and create visually cohesive documents
* **Flexible export** Export as Word or PDF, open in a separate browser, or export as Markdown from Markdown-format documents
**Learn more about AI Docs:** [https://www.genspark.ai/blog/genspark-ai-docs](https://www.genspark.ai/blog/genspark-ai-docs)
## Quick Start {#quick-start data-section-type=overview}
Get started in four simple steps:
1. **Go to AI Docs**
From the Genspark homepage, click "AI Docs" at the bottom, or select "AI Docs" from the "New" menu in the sidebar.
2. **Choose your format**
Pick the format that fits the document — **Rich Text**, **Markdown**, **Web Doc**, or **Word**. The choice is permanent: the format is locked once the document has content. Want manual editing later? Rich Text and Markdown are the editable formats — Web Doc and Word are preview-only.
3. **Describe your needs**
Tell AI what you want to create a resume, report, contract, or any document type. Here are some sample prompts:
```trythisprompt
Create a professional resume for a software engineer position based on my LinkedIn profile
```
```trythisprompt
Design a menu for a boutique coffee shop with coffee, tea, and dessert categories
```
```trythisprompt
Convert this handwritten class schedule to a formatted electronic document
```
4. **Get results**
AI Docs automatically generates your document with proper structure and formatting. You can then refine through conversation, make manual edits, and export your document. The system automatically creates "Save Points" for version tracking.
## Feature & How-To Guides {#feature-guide data-section-type=feature}
### AI Edit {#AI-Edit}
Select any text for precise AI editing without regenerating the entire document.
1. Select text with your mouse — the selected content appears above the chat box
2. Enter your instruction, such as:
* "Rewrite this in a more professional tone"
* "Expand this section with specific examples"
* "Translate this paragraph to Spanish"
* "Make this more concise"
3. Apply to let AI rewrite the selected content
In **Rich Text**, you can also select text and use the in-place **AI Rewrite** control, or right-click and choose **AI Tools** → **AI Write**. **Markdown** has no in-place AI tools — selecting text there sends it to the chat bar as context instead.
**AI Rewrite** and **AI Write** send your entire document as context — a one-line edit bills input tokens for the whole doc. Manual editing bills nothing.
### Manual Edit {#Manual-Edit}
You can make direct edits in the **Rich Text** or **Markdown** editor to adjust text formatting, layout, and insert elements like links, images, and tables.
Which formats are editable:
| Format | Manual editing |
| --- | --- |
| **Rich Text** | Yes — full visual editor |
| **Markdown** | Yes — plain-text editor |
| **Web Doc** | No — preview only |
| **Word** | No — preview only |
The output format is chosen up front and is **locked once the document has content**. The format selector works before the document exists; after that it greys out — if you try to switch, it reads: "To use a different format, please start a new project."
There's no Save button — your edits save automatically (about a second after you stop typing), and they're included when you export, even if you export right after editing.
### Edit any document {#edit-any-document}
Web Doc and Word documents can't be hand-edited, but you can still change them:
1. **Open the document.**
2. **Highlight the text** you want to change — this works in every format. In **Rich Text** and **Word**, you can also **click a generated element** (a table, for example) to target it; **Web Doc** is text-selection only.
3. **Prompt the change in the chat bar** on the left.
Selection-targeted editing works this way for every format — including preview-only Web Doc and Word documents.
### Export {#export}
Click **View & Export** in the top right corner of the document. The button appears once your document has content and is disabled while the agent is still working. The menu depends on your format:
* **Rich Text / Web Doc** — Export as Word Document · Export as PDF · Open in separate browser · Make a Copy
* **Markdown** — the same, plus Export as Markdown
* **Word** — Download Word Document · Export as PDF · Make a Copy
**Open in separate browser** opens your document as a web page in a new tab — that's the HTML view, not a file download. **Make a Copy** duplicates the document into a new project for independent editing.
Certain export formats (Word, PDF, Markdown) are included with **Plus, Pro, and Team** plans.
### Version Management {#Version-Management}
AI Docs uses a Save Point system to track every modification:
* **AI edits:** Each AI prompt creates a new Save Point (none is created if the content didn't change)
* **Manual edits:** A manual edit creates a Save Point named **"[Manual Edit]"**; consecutive manual changes by the same person are merged into that same version — a second editor or a version conflict starts a new Save Point
* **Complete tracking:** Every Save Point captures the full document state and modification information
**Here's how to use Save Points:**
1. Click version dropdown in upper left corner of editor
2. View all historical versions (reverse chronological order)
3. Select any version to preview its content
4. Click "Rollback to" to restore your document to that version
The Save Point dropdown appears only once there's more than one Save Point, and it's hidden on mobile and in read-only documents.
**Export follows the version you're previewing.** If you're viewing an older Save Point, that older version is what exports — not your latest edits. Make sure you're on the version you want before you export.
**Here are some common scenarios:**
* **Recover deleted content** Find the Save Point before deletion, preview to confirm, then rollback
* **Compare different approaches** Try version B with AI, compare it with version A, rollback if needed, or try version C
* **Track document evolution** Review how your document was refined step by step through each Save Point
### Format Conversion {#Format-Conversion}
The format is chosen when you create the document and is **locked once the document has content** — the format selector greys out, and if you try to switch you'll see: "To use a different format, please start a new project." There's no format conversion after the document is generated; to use another format, start a new project and choose it at creation.
### Template Library {#template-library}
We provide hundreds of professional templates covering job applications, business communications, reports and analysis, legal and finance, creative and marketing, event planning, and more.
You can select a template on the AI Docs homepage, enter your prompt to customize content, or use "Manual Edit" or "AI Edit" to refine your document.
* **Share:** Click "Share" in the top right corner, adjust permission settings, and share your project link
* **Export:** Click **View & Export** in the top right corner and choose your format — the options depend on the document's format (Markdown documents also offer Export as Markdown)
## FAQ {#faq data-section-type=faq}
### Should I choose Rich Text or Markdown? {#rich-text-vs-markdown}
Choose "Rich Text" if you need beautiful layout and visual editing; choose "Markdown" if you focus on content, need version control, or write technical documentation.
### Is AI-generated content accurate? {#content-accuracy}
AI uses the latest large model technology with high accuracy, but we recommend human review of key information.
### How long can a document be generated at once? {#document-length-limit}
There's no strict limit, but we recommend generating complex documents in segments for better quality control.
### Can I recover my current work after rolling back to a previous version? {#recover-after-rollback}
No, when you rollback to a previous version, all changes made after that version will be lost and cannot be recovered — and the conversation history after that point is cleared too. Make sure to review the version carefully before confirming the rollback.
### How to delete unwanted Save Points? {#delete-save-points}
Currently the system automatically manages all historical records to ensure complete version tracking.
# Genspark Code {#genspark-code data-doc-group=build-suite}
Source: https://www.genspark.ai/helpcenter/ai-developer
## What is Genspark Code? {#what-is-ai-developer data-section-type=overview}
**Genspark Code** (formerly AI Developer) is a L4 autonomous AI coding agent that enables everyone to build complete applications without coding skills. Evolving from traditional AI copilots (L3 tools like Cursor or Lovable), it marks the leap to L4 full autonomy, where AI independently plans, codes, tests, and deploys. This is the future of vibe working — AI handles technical execution so humans can focus on creativity, strategy, and innovation.
**Learn more about Genspark Code:** https://www.genspark.ai/blog/genspark-ai-developer
### Projects tab {#projects-tab}
The Genspark Code home page has a **Projects** tab in the category row. It shows the projects you've built with Genspark Code so you can pick work back up where you left off.
* **Reopen in one click** — each project card is a direct link into the project; no intermediate step.
* **Pinned projects first** — pinned projects stay at the top of the list across pages; each card shows a cover thumbnail and when it was last edited.
* **A paginated list** — there's no search box or filter; longer lists load in pages with a **Load more** button.
* **Genspark Code projects only** — the tab lists code projects (deleted ones excluded). It's not a general history, so it won't show AI Slides, AI Docs, or Super Agent work.
The projects tab isn't the page's default view — the Genspark Code home page still opens with **Web** selected. Switch to **Projects** when you want to get back to your work.
When there are no projects yet, the tab shows **"No projects yet"** with a **Start building** button. A project that has no name appears as **Untitled project**; if a list fails to load, you'll see a **Retry** option.
**Don't confuse this tab with the recent-projects drawer:** the Genspark Code page also has a pre-existing recent-projects drawer — a slide-out from the top-left that opens your project list. That drawer is a separate, older feature; it isn't the Projects tab and still works alongside it.
## Tutorials {#tutorials data-section-type=feature}
Genspark Code builds in the five categories **Web**, **Mobile Apps**, **Game**, **E-commerce**, and **Existing code**. Pick one on the home page to start a project, then open the **Tutorial** tab in the panel on the right for a step-by-step guide.
For the complete tutorials, visit https://www.genspark.ai/code/.
## FAQ {#faq data-section-type=faq}
### What is the Projects tab in Genspark Code? {#what-is-projects-tab}
The Projects tab lists the projects you've built with Genspark Code so you can reopen them — each card links straight into the project. It lists code projects only, and it isn't the page's default view: the home page opens on **Web**; switch to **Projects** to see your work.
### Why does the Genspark Code home page open on Web instead of Projects? {#why-home-page-opens-on-web}
The Projects tab is an entry point, not the default — **Web** is selected when the page loads. Click **Projects** in the category row to see your projects.
# AI Meeting Notes {#ai-meeting-notes data-doc-group=tools}
Source: https://www.genspark.ai/helpcenter/ai-meeting-notes
## What is AI Meeting Notes? {#what-is-ai-meeting-notes data-section-type=overview}
**Genspark AI Meeting Notes** is the world's first AI meeting notes tool for Apple Watch, also available on iOS, Android, and Web. Just double-tap on your Apple Watch or tap once on your phone to instantly turn any meeting into comprehensive, professional meeting notes.
**Learn more about AI Meeting Notes:** https://www.genspark.ai/blog/genspark-ai-meeting-notes
## Quick Start {#quick-start data-section-type=overview}
1. Open the Genspark app and go to **AI Meeting Notes**
2. Tap **Connect Calendar** to link your Google Calendar or Outlook — this lets the app show your upcoming meetings and enables Meeting Bot to join them automatically
3. Tap **+ New** to start recording. When the meeting ends, tap **End** to generate your notes
4. The system automatically analyzes the recording and produces meeting notes with a full transcript. You can then ask AI questions about the meeting content
AI Meeting Notes is available on web, mobile app, and Apple Watch.
## Feature & How-To Guides {#feature-guide data-section-type=feature}
### Meeting Recording {#meeting-recording}
- **Language support**: All major languages are supported
- **Recording length**: No time limit
- **Audio source**: Records via your device's microphone only (system audio is not supported)
- **Recording interruptions**: Recording auto-resumes if interrupted. If the app closes, reopen it or tap Resume to pick up where you left off
- **Generating notes**: Tap **End** to upload. Keep the app open until progress passes 50% to avoid upload failures
- **If generation fails**: On web, click the Resume button on the recording entry to re-upload. On mobile, reopen the app to automatically retry. On Apple Watch, swipe left on the entry and tap the phone icon to regenerate
- **Import audio**: Available on the mobile app and web
- **Live notes**: During recording, you can jot down notes in the notes input box at any time — they'll be used as additional context when generating your meeting summary
### Meeting Notes Management {#meeting-notes-management}
- **Edit notes**: Click anywhere in the meeting notes area to edit (undo and version history are not supported in the current version)
- **Copy transcript**: Transcripts can't be edited directly, but you can copy the full content with one click
- **Delete a recording**: On mobile, swipe left on the entry and select Delete. On desktop, click the three-dot menu (⋯) to the right of the entry and select Delete
- **Chat**: Ask AI questions about your meeting content. Responses appear in the chat stream and won't overwrite your original meeting notes
- **Search**: The search bar on the main page searches across all meeting titles and summaries. Inside a transcript, you can search for specific keywords and jump to where they appear
### Sharing & Export {#sharing-export}
- **Share with participants**: After the meeting ends, expand the **Participants** section on the meeting notes detail page and click **Share with X participants** — the system will automatically email the notes to all attendees
- **Export to Google Docs**: send notes or full transcript to a new Google Doc; ready-to-open link; one-time auth scoped only to files it creates
- **Email sharing**: You can also type "send email to [XX@example.com](mailto:XX@example.com)" in the chat box, and the system will send the notes via the Genspark team email ([team@genspark.ai](mailto:team@genspark.ai))
- **Share link**: Generate a shareable link — set General Access to "Anyone with the link," or invite specific team members by email, then copy and share the link
- **Save as PDF**: Export your notes as a PDF, great for printing or archiving
- **Save to Notion**: If you have Notion installed, you can save your meeting notes to Notion with one click
### Apple Watch {#apple-watch-usage}
- **System requirements**: watchOS 10.0 or later
- **Installation**: After installing the Genspark app on your iPhone, the Watch app will automatically install on your Apple Watch. If it doesn't, manually install it from the Available Apps section in the Watch app on your iPhone
- **Recording a meeting**: Open the Genspark app on your Apple Watch and tap to start recording. You can switch apps or turn off the screen during recording — it keeps going in the background. When you're done, tap to stop, and the recording will automatically sync to your iPhone and the web
### Meeting Bot & Calendar Integration {#meeting-bot}
Meeting Bot is an AI assistant that automatically joins your online meetings, records them, and generates meeting notes when they're done — no manual steps required.
**Step 1 — Connect your calendar**
Tap **Connect Calendar** on the AI Meeting Notes home screen to link your Google Calendar or Outlook (if you haven't done this already). Once connected, the **Coming up** section will show your public meetings for the next three days (private events are excluded).
**Step 2 — Enable Auto-join**
1. Click the settings button in the top-right corner, go to **Settings → AI Meeting Notes**
2. Under **Meeting Bot**, turn on **Auto-join**
With Auto-join on, Meeting Bot will automatically join and record all calendar meetings where a meeting link is detected. You can also start it manually for any specific meeting — just tap it in the **Coming up** section and hit **Join Meeting**.
**Additional settings**:
- **Bot name**: Customize the name shown in the meeting room — defaults to "Genspark Meeting Bot"
**Supported platforms**: Zoom, Google Meet, Microsoft Teams, Webex, GoToMeeting
**Credits**: Recording with Meeting Bot uses Credits, billed by actual recording time.
*Note: Meeting Bot can only join meetings where a meeting link is included in the calendar invite. If there's no link in the invite, the Bot won't be able to join automatically.*
### Auto-share {#auto-share}
When enabled, Auto-share automatically emails the meeting notes to all participants as soon as the meeting ends — no manual steps needed.
To turn it on: go to **Settings → AI Meeting Notes** and enable **Auto-share**.
### Language Settings {#language-settings}
There are two separate language settings in AI Meeting Notes, and they work independently:
**1. Transcript language**
The transcript is always auto-detected based on what was actually spoken. It accurately reflects the real speech — even in multilingual meetings where participants switch between languages, the transcript will capture each language as heard.
**2. Meeting notes language**
The language your final notes are generated in is controlled separately, via **Settings → AI Meeting Notes → Notes Language**:
- **Auto detect** (default): The AI reads the transcript and generates notes in the transcript's dominant language. You don't need to configure anything — it works automatically.
- **Manual selection**: If you choose a specific language, notes will always be generated in that language regardless of what language the meeting was held in.
## FAQ {#faq data-section-type=faq}
### Why is my recording taking a long time to process? {#recording-processing-time}
**Possible causes**:
- Unstable network connection during upload
- The app was closed before the upload completed
- The server is handling a high volume of requests
**What to do**:
- **To prevent this**: Make sure you have a stable connection before uploading, and keep the app open until progress passes 50%
- **If it already failed**: Check your connection and reopen Meeting Notes to trigger a re-upload
### What should I do if there's no sound after recording on web? {#no-sound-web}
**Cause**:
- The wrong microphone device was selected
**What to do**:
- Make sure you've selected the correct microphone before you start recording
- Check the audio waveform right after starting — it should show activity
- If there's no waveform, stop the recording, switch your microphone in the browser settings, and try again
### Can I download the original audio file? {#download-audio}
To protect user privacy, Genspark does not save or provide access to original audio files. Only the transcript text and meeting notes are available.
### Why didn't Meeting Bot join my meeting? {#meeting-bot-not-joining}
Check the following:
- Is your calendar connected (Google Calendar or Outlook)?
- Is **Auto-join** turned on in Settings?
- Does the calendar invite include a meeting link (Zoom, Google Meet, Teams, etc.)?
- Do you have enough Credits in your account?
If everything looks good, try disconnecting and reconnecting your calendar.
### My "Coming up" meetings disappeared and Meeting Bot stopped joining automatically. What happened? {#coming-up-missing}
Your calendar connection has likely expired. Calendar integrations, especially Outlook, use time-limited access tokens. Once a token expires, Genspark can no longer read your calendar, so the **Coming up** section goes blank and Meeting Bot stops joining automatically.
To fix it, go to **AI Meeting Notes → Connect Calendar**, disconnect your current calendar, and reconnect it. After reconnecting, your upcoming meetings should reappear in the **Coming up** section and Meeting Bot will resume auto-joining.
# AI Pods {#ai-pods data-doc-group=content-creation}
Source: https://www.genspark.ai/helpcenter/ai-pods
## What is AI Pods? {#what-is-ai-pods data-section-type=overview}
**AI Pods** is Genspark's AI podcast generator. Describe what you want in a single prompt, and it turns a topic, document, or link into a podcast episode — researching the content, writing the script, creating the hosts' dialogue, and producing the audio for you.
**Learn more about AI Pods:** [https://www.genspark.ai/blog/genspark-ai-pods](https://www.genspark.ai/blog/genspark-ai-pods)
## Quick Start {#quick-start data-section-type=overview}
Get started in three steps:
1. **Go to AI Pods:**
Select "AI Pods" from the "New" menu in the sidebar, or click "All Agents" on the Genspark homepage and find "AI Pods".
2. **Tell AI Pods what you want to create:**
Sample prompts:
```trythisprompt
Give me a daily news briefing podcast on today's tech headlines
```
```trythisprompt
Create a podcast explaining the research paper I uploaded
```
```trythisprompt
Make a workout motivation podcast with upbeat music and commentary
```
3. **Get your podcast:**
AI Pods usually generates your episode in 5–15 minutes. Listen to it directly, share the project link, or use **Export Audio** to save it to AI Drive and download it from there.
## Feature & How-To Guides {#feature-guide data-section-type=feature}
### Podcast Creation {#podcast-creation}
Enter a prompt to create your podcast. AI Pods writes the script and produces the audio automatically — the language, the number of hosts, the length, the style, and whether you want audio or a video podcast are all set in the prompt.
* **Style:** Describe the tone you want — conversational, educational, entertaining, or professional.
* **Editing:** After creation, you can regenerate the cover image. To change anything else, create a new podcast with an updated prompt.
### What You Can Give It {#sources-and-inputs}
AI Pods works from what you put in your prompt:
* **Upload a document** — PDF, DOCX, TXT, and similar document files. AI Pods reads the document and builds the episode around it.
* **Paste your text** — drop in notes, an article, or any block of text.
* **Describe a topic** — give it a subject and AI Pods researches it and writes the episode.
* **Give it a link** — point it at a URL and AI Pods reads the page and turns it into an episode.
### Episode Length {#episode-length}
You can ask for a duration in your prompt — for example, "make it about 10 minutes".
* The duration is a **target**, not a fixed setting. AI Pods aims the script at the length you ask for, so the finished episode may be a little longer or shorter.
* Generating an episode usually takes 5–15 minutes.
### Speakers {#speakers}
* Episodes use **2 hosts** by default, and you can ask for a single host in your prompt. Auto-generated audio supports up to 2 speakers.
* If you ask for **3 or more speakers**, AI Pods produces a **script without audio**, so you can record the dialogue yourself.
### Language {#language}
The spoken language comes from your prompt — write the language you want the episode in, for example "in Japanese" or "en español". There is no separate language setting.
* **Supported languages:** 24 languages, including English, Chinese, Arabic, German, Spanish, French, Hindi, Indonesian, Italian, Japanese, Korean, Portuguese, Russian, Dutch, Polish, Thai, Turkish, Vietnamese, Romanian, Ukrainian, Bengali, Marathi, Tamil, and Telugu.
* Name the language explicitly, especially when the topic doesn't make it obvious.
### Video Podcast {#video-podcast}
AI Pods can also produce a **video podcast**: the same host dialogue paired with visual slides and reference images, plus an automatically generated cover image. Use it when you want something to watch, post, or present rather than only listen to.
* Ask for a video podcast in your prompt.
* Export the video with the same export button — it reads **Export Video** when your episode has video.
### Publish, Share & Export {#publish-share-export}
Once your podcast is ready, you can publish it, share it, or export it.
**Publish your podcast:**
* Click **Publish** to share your episode with the Genspark community. You need to be signed in, and publishing is available once the episode has audio. Published episodes can appear on the AI Pods page.
* Publishing takes effect right away — there is no review step.
* To take your episode down, click the **Published** button to unpublish. It is removed from the published listing; the project stays in your account.
**Share your podcast:**
* For private sharing, click the **Share** button in the upper right corner, adjust the permission settings, and share the project link with others.
**Export your podcast:**
* **Export** — with an audio episode the button reads **Export Audio**; it saves your episode to AI Drive, where you download the audio file. If your episode has video, the same button reads **Export Video** and saves both the audio and the video.
## Credits & Cost {#credits-cost data-section-type=feature}
AI Pods episodes are billed in credits.
* **Credits, not Free Quota.** Audio generation is always billed in credits, even on plans that include Free Quota. See [Membership Plans](https://www.genspark.ai/helpcenter/membership-plans) for what Free Quota covers.
* **Cost tracks the script.** The longer the episode's script, the more credits it uses, so shorter episodes cost less.
* **Failed audio runs aren't billed.** If the audio generation fails, you aren't charged for it. You can ask again in the same project, or start a new episode with a shorter or clearer prompt.
For how credits are counted and charged, see the [Credits Guide](https://www.genspark.ai/helpcenter/credits-guide).
## Troubleshooting {#troubleshooting data-section-type=troubleshooting}
### My episode has no audio at all. {#no-audio}
A few things to check:
* Open the project and check whether generation has finished. If it is still working, give it up to 15 minutes.
* Did audio generation fail? If so, you aren't charged for it — ask again in the same project, or start a new episode with a shorter or clearer prompt.
* Did you ask for 3 or more speakers? See [I asked for 3 or more speakers and got a script instead of audio](#three-speakers).
### I asked for 3 or more speakers and got a script instead of audio. {#faq-three-speakers}
Auto-generated audio supports up to 2 speakers, so a request for 3 or more returns a script without audio for you to record yourself. Ask for 1 or 2 speakers in the prompt to get an audio episode.
### A run failed. What happens to my credits? {#failed-run}
If the audio generation failed, you aren't charged for it. Ask again in the same project, or start a new episode with a clearer prompt. For how credits are counted and charged, see the [Credits Guide](https://www.genspark.ai/helpcenter/credits-guide).
### The episode came out in the wrong language. {#language}
The spoken language follows your prompt, and there is no language setting to check. Create a new episode and state the language explicitly in your prompt — for example, "Create a 5-minute podcast in Japanese about…". See [Language](#language).
## FAQ {#faq data-section-type=faq}
### How do I report a problem? {#report}
Share the project with our support team, then contact support with the details:
1. Open the project and click **Share**.
2. Invite **support@genspark.ai** by email as a **Viewer**.
3. Email **support@genspark.ai** with the project link, what went wrong, and your approximate credit balance before and after.
# Call for me {#call-for-me data-doc-group=tools}
Source: https://www.genspark.ai/helpcenter/call-for-me
## What is Call for Me? {#what-is-call-for-me data-section-type=overview}
Call for Me is Genspark's AI calling agent. From a single prompt, it places a real phone call on your behalf, handles the full conversation in a natural voice, navigates automated menus, and reports back with a transcript and summary. Use it to book reservations, confirm appointments, check hours, break language barriers, and more.
**Learn more about Call for Me**: https://www.genspark.ai/blog/genspark-ai-calls-personal-phones
## Quick Start {#quick-start data-section-type=overview}
**Verify your phone number**
- Select your country code, enter your phone number, and verify it with the code sent to you by text message. Voice-call verification is a separate, optional step that enables your number to appear as the caller ID.
**Choose who to call**
- Pick **Place** to search for a business by name or type (the number and address come from Google Maps automatically), or **Contact** to enter a name and number directly.
**Describe the call purpose**
- Tell the AI what to accomplish, including any key details and fallbacks.
**Place or schedule the call**
- Choose **Call Now** or **Schedule Call**, then confirm. The AI makes the call, and you can review it afterward.
**Example Prompt**
> "Call Rosa's Trattoria and book a table for 4 tonight at 7 pm. Ask if they have outdoor seating. If 7 isn't available, try 7:30."
**What this produces:** The AI calls the restaurant, requests the booking with your preferences, handles the back-and-forth, and returns a summary of what was confirmed.
---
## Feature & How-To Guides {#features-and-how-to-guides data-section-type=feature}
### Choosing a Voice {#choosing-a-voice}
Call for Me offers 10 AI voices — 5 female (Mira, Alina, Sophia, Ruby, Willow) and 5 male (Cedric, Finn, Ryan, Mason, Leo). The default is Mira.
**Open Call for Me settings**
- Find Call for Me in your agent list and click its gear icon.
**Pick a voice**
- Choose any voice from the list, and click the play button next to a voice to hear a preview.
**Save**
- Your selection is saved and used for future calls until you change it.
### Scheduling a Call {#scheduling-a-call}
When you set up a call, choose **Schedule Call** instead of **Call Now** and pick the date and time. Call for Me places the call then and posts the transcript and summary to your history, just like an immediate call.
**Tips:**
- To cancel a scheduled call, cancel it where you created it: calls scheduled with the Schedule Call button appear in **Task List** and **History** — open the call there and cancel it before its scheduled time. If you scheduled the call by asking in a conversation, just ask in that conversation to cancel it.
### Viewing Call Results and Transcripts {#viewing-call-results-and-transcripts}
After a call finishes, open **Task List** or **History** to see a full text transcript and a short summary; a recording of the call may also be available for playback. You will also receive a task-finish email with the call summary at your account email address.
**Tips:**
- Use the **Docs** tool in Super Agent to turn a transcript into a clean, structured document for records or sharing.
- If the recipient doesn't answer, the call is marked unsuccessful — you can redial or try another contact.
- To cancel a call in progress, cancel it before it connects; once connected, it's best to let the AI finish naturally.
### Advanced: Batch Calls, Hub, and Custom Assistants {#advanced-batch-calls-hub-and-custom-assistants}
- **Batch calls** — use Super Agent to run several calls together (for example, ring three restaurants to compare availability, or notify a list of people) instead of one at a time.
- **Hub project management** — keep related calls in one Hub: group call tasks, maintain a contact library, and run follow-ups — useful for repeat outreach.
- **Custom Call for Me Assistant** — build a Custom Super Agent for scenarios you handle often, with the situation, style, and details it should remember (for example, a Restaurant Reservation Assistant that knows your usual party size, seating, and dietary needs).
---
## Pricing and Credits {#pricing-and-credits data-section-type=feature}
> **Paid plan required:** Phone number registration requires Plus or higher. For Call for Me, any paid plan can verify up to 3 phone numbers.
Call for Me runs on credits, and your phone bill is never charged — calls go out over Genspark's lines, not yours.
- Each call costs a flat **500 credits** for the real-time voice model.
- A few additional credits cover analyzing the call and writing the summary, at standard text-model rates.
---
## FAQ {#faq data-section-type=faq}
### Whose number shows on caller ID? {#caller-id}
- It depends on how you verified. You can verify up to 3 phone numbers, set a default, and choose which verified number to show for each call. If you verified your number by voice call and the recipient's network supports it, your registered number is shown; otherwise the call falls back to a virtual number.
### What languages does it support? {#languages-supported}
- Call for Me can handle calls in multiple languages and translate in real time during a call — useful when you and the business speak different languages. Translation isn't flawless, so keep that in mind for high-stakes calls.
### Can it get through automated phone menus? {#automated-menus}
- Yes. The AI can press the right number keys to navigate "press 1 for…" style menus and reach the right option or a person.
### Do failed calls still cost credits? {#failed-calls-credits}
- You're charged only when a call connects. If nobody picks up, the line is busy, or the call can't connect, the 500 credits are refunded automatically. A call answered by voicemail counts as connected and is charged, as is a call that connects and then drops mid-conversation.
### What happens if no one answers? {#no-answer}
- The call is marked unsuccessful and nothing further happens automatically — you can redial or use another contact. If the call fails to connect at dial time because of certain connection issues, Call for Me may automatically retry using a system number. It does not automatically redial after a no-answer or a mid-call drop.
### How long can a call last? {#call-duration}
- Most calls finish within a few minutes; a call can run up to about an hour, subject to carrier network limits.
### How can I remove my phone number? {#remove-phone-number}
- To remove your phone number, open Call for Me, go to **Settings**, and delete the phone number.
# Custom Agent {#custom-agent data-doc-group=custom-agents}
Source: https://www.genspark.ai/helpcenter/custom-super-agent
## What is Custom Agent? {#what-is-custom-agent data-section-type=overview}
**Genspark Custom Agent** is the world's first one-prompt AI agent creation platform. Simply describe what you need — industry reports, stock analysis, or personalized content — and your agent is ready instantly. Build in minutes without coding, reuse effortlessly by chatting or @mentioning, share globally through the Custom Agent Store, and discover amazing agents built by the community.
**Learn more about Custom Agent**: https://www.genspark.ai/blog/genspark-custom-agent
## Creation {#creation data-section-type=feature}
### Entry Points {#entry-points}
- **Homepage Entry**: Click "Custom Agent" on the Genspark homepage
- **Sidebar Quick Access**: Click the "+ New" button in the left sidebar—Custom Agent Store appears at the top of the list
- **In-Conversation Invocation**: Type @ in the Super Agent chat window to invoke
### Creation Process {#creation-process}
1. Click the "Create New" button in the upper right corner of the Custom Agent Store page
2. Create using a structured template: [Role Definition] + [Core Functions] + [Output Standards]
**Example Prompt:**
```prompt
Create an AI industry analyst that generates weekly reports on funding trends, technology developments, and policy interpretations for investment decision-making
```
3. Test the agent in the Preview window, adjust instructions based on test results, and click "Update" to save changes
## Usage {#usage data-section-type=feature}
### Managing Custom Agents {#managing-custom-agents}
- **Edit**: Click the three-dot menu on the agent card, select Configure, modify settings through conversational interaction or manual edits, then click Update to save
- **Delete**: Click the three-dot menu, select Delete and confirm (deletion is permanent and irreversible)
- **Share**: Click the three-dot menu, select Configure, then click the Share button to choose sharing scope:
- Only me
- Anyone with the link
- Custom Agent Store
### Custom Agent Store {#custom-agent-store}
- **My Super Agents**: Agents you've created and favorited
- **Trending**: Sorted by number of favorites
- **By Genspark**: Officially published agents
- **Category Filters**: Writing / Productivity / Research & Analysis / Education / Lifestyle / Others
**How to Use:**
- Click an agent card to view details and start a conversation
- Mark as favorite to add to "My Super Agents"—invoke via @ in Super Agent
- Publish an agent: Click Share on the Configure page and select "Custom Agent Store"
## Pro Tips {#pro-tips data-section-type=feature}
### Invoking Custom Agents Within Super Agent {#invoking-agents}
Type @ in a conversation to invoke Custom Agents. You can combine multiple agents to collaboratively complete complex tasks.
For example, when drafting an investment report, you can sequentially invoke a Financial Analysis Custom Agent and an Industry Research Custom Agent, then have Super Agent synthesize a complete report.
### Optimizing Instructions Through Conversation {#optimizing-instructions}
Describe your requirements in natural language in the Create page dialog—the system will automatically update the Instructions.
**Key Elements**:
- **Role Definition**: Clearly specify professional identity
- **Output Format**: Specify presentation style (e.g., tables, sections)
- **Behavioral Boundaries**: Set response scope
Test in Preview after each modification and continuously refine until satisfied.
### Leveraging Conversation Starters to Guide Users {#conversation-starters}
Conversation Starters help users understand how to interact with your agent.
Adjust using prompts on the Create page, or manually edit the Conversation Starters text box at the bottom of the left panel on the Configure page.
## FAQ {#faq data-section-type=faq}
### How can I make the agent's responses more accurate? {#accurate-responses}
Use structured prompts to clearly define roles and tasks, enhance Instructions with detailed guidance, test in Preview, and continuously optimize.
### Can a deleted agent be recovered? {#deleted-agent-recovery}
No, deletion is permanent. We recommend exporting the conversation history from agent creation before deleting.
### Can I edit agents from the Agent Store? {#edit-store-agents}
No, favorited agents can only be used—you cannot view or edit their instructions.
### If I use someone else's shared Custom Agent, will my chat history be leaked? {#chat-history-privacy}
No. Each user's conversation history is completely isolated, and creators cannot view other users' conversations. Sharing only includes basic information like name and description—Instructions and conversation data are not included.
# Hub {#hub data-doc-group=custom-agents}
Source: https://www.genspark.ai/helpcenter/hub
## What is Hub? {#what-is-hub data-section-type=overview}
**Genspark Hub** is a centralized workspace for managing your projects with shared context. Upload files once and they're available across all conversations. Set custom instructions to guide how Genspark responds. Organize related projects together—no need to repeat context or re-upload files. Ideal for ongoing work like research, content creation, product development, and collaboration.
**Work with your team**: Invite members to your Hub. Everyone can access the same files, reference previous conversations, and pick up work on any project seamlessly.
**Learn more about Hub:** https://www.genspark.ai/blog/genspark-hub
## Quick Start {#quick-start data-section-type=overview}
1. **Create and Access Your Hub:**
- Access the Hub Interface: Click "Hub" in the left sidebar or visit https://www.genspark.ai/hub
- Create a New Hub: Click "+ New", enter a name, theme color, and hub description, then click "Create Hub"
- Access Your Hub: Find and click your hub in the "Your Hubs" list
2. **Start a Project in Your Hub:**
- Type your prompt to begin. Super Agent is the default, or use @ to select other agents: AI Slides, AI Sheets, AI Docs, Genspark Code, AI Designer, Clip Genius, AI Pods, My Custom Super Agent
- Access all your projects in "Recent Tasks" on the Hub main page to view history or continue work
## Feature & How-To Guides {#feature-guide data-section-type=feature}
### Projects in Hub {#projects-in-hub}
You can find all projects within this Hub in "Recent Tasks" on the main page
#### Cross-Project Context {#cross-project-context}
Projects within the same Hub share conversation history—reference any previous project in this Hub, and Genspark automatically retrieves and applies relevant history
Use Cases:
- **Iterative research**: "Generate a competitor analysis based on last week's market research project"
- **Meeting follow-up**: "Draft an action plan referencing yesterday's meeting notes project"
- **Continuous updates**: "Update Q4 forecast based on Q3 analysis project conclusions"
- **Knowledge reuse**: "Design new product features using insights from customer interview project"
#### Project Management {#project-management}
Organize projects across Hubs in three ways:
1. **From within a Project:** Click "⋯" in the top right and select "Add to Hub", "Change Hub", or "Remove from Hub"
2. **From Task List:** Open Task List in the top left, click "⋯" on the project to manage its Hub assignment
3. **From Hub main page:** Hover over a project in "Recent Tasks", click "⋯" to move it to another Hub or remove it
**Note**: Each project can belong to only one Hub at a time
### Hub Files {#hub-files}
Click "Files" to add files from your computer or AI Drive
- **Supported**: Documents (doc, docx, pdf, md), code (html, py, json), presentations (ppt, pptx), spreadsheets (xlsx, csv), images (png, jpg, jpeg, webp)
- **Not supported**: Audio, video, compressed files
- **Limits**: 100 files per Hub, 10MB per file
Hub automatically parses file content and references it during conversations
**Best Practices**
**1. Use Descriptive Filenames:** brand_guidelines_2024.pdf instead of document1.pdf
**2. Keep Files Minimal:** Too many files increase credit consumption
**3. Specify Files in Prompts:** "Generate a report based on Q4_sales_data.xlsx"
### Custom Instructions {#custom-instructions}
Set work instructions that apply to all projects in the Hub:
- Specify focus topics or domains
- Define tone, style, or response format
- Explain team terminology or company concepts
- Set file usage priority
**Best Practices:**
1. **Test Before Finalizing** — Verify with prompts first
2. **List Non-Negotiable Rules** — E.g., "All marketing copy must include a disclaimer"
3. **Provide Style Examples** — Show ideal output format
4. **Define Proprietary Terms** — E.g., "Sparkle = Q4 product launch codename"
### Members {#members}
Hub members can view all files, access all project conversation history, continue work on any project, and create new projects with shared context
#### Invite Members (Hub Creator Only) {#invite-members}
Hub creators can invite members in the following ways:
- **From Hub main page**: Click "Members" > "Invite" and enter email address
- **From a Project**: Invite directly from any project within the Hub
- **Move a Project with members**: When a project with members is moved into the Hub, those members automatically gain access to the Hub
#### Remove Members (Hub Creator Only) {#remove-members}
Remove a member from the Hub main page: Click "Members" > trash icon next to the member > "Remove"
#### Leave a Hub {#leave-hub}
Members can leave a Hub in two ways:
- **From Your Hubs page**: Hover over the Hub, click "⋯" > "Leave Hub"
- **From inside the Hub**: Click "⋯" in the top right > "Edit Hub" > "Leave Hub"
**Note**: Hub owners cannot leave a Hub—only delete it (see Delete Hub section)
### Edit Hub {#edit-hub}
Access the edit page in two ways:
- **From Your Hubs page**: Hover over the Hub, click "⋯" > "Edit Hub"
- **From inside the Hub**: Click "⋯" in the top right > "Edit Hub"
Editable content: Hub name, theme color, Hub description
### Delete Hub {#delete-hub}
Deleting a Hub removes all files and projects, this action cannot be undone
## FAQ {#faq data-section-type=faq}
### Why is the Hub file size limit 10MB? What if my file is too large to upload? {#file-size-limit}
The 10MB limit helps you save credits and ensures better results. Large files consume excessive credits and often produce poor-quality analysis.
If your file is too large:
- Extract only the relevant sections
- Summarize the content before uploading
- Split into smaller focused files
### How many projects can a Hub contain? {#hub-project-limit}
There's no strict limit on the number of projects per Hub. Performance depends on search results, not total project count.
### Can I download or export files from a Hub? {#download-hub-files}
Yes. Click "Files" on the Hub homepage, hover over the file you want, and click to download it.
### Is there a time limit on Hub file storage? {#hub-storage-time}
No. Hub provides persistent file storage with no time limit. Files remain accessible indefinitely. Limits are 100 files per Hub and 10MB per file.
# Genspark Claw {#genspark-claw data-doc-group=ai-team}
Source: https://www.genspark.ai/helpcenter/genspark-claw
## What is Genspark Claw? {#what-is-genspark-claw data-section-type=overview}
**Genspark Claw** is your first AI employee — a personal assistant that works on your behalf, whether running on a dedicated cloud computer or locally on your own machine. Unlike chat-based AI tools, Claw connects to your apps, remembers your context, and can take action for you across email, calendar, messaging, and more.
- **Flexible deployment** — Run Claw on a cloud computer (always on, 24/7) or locally via the desktop app (no subscription required, uses your own machine)
- **Works where you message** — Chat with Claw in WhatsApp, Slack, Teams, Telegram, LINE, Discord, Signal, Google Chat, Feishu, and more
- **30+ built-in skills** — Research, email, calendar, slides, docs, code, image generation, video generation, phone calls, and more, all pre-installed
- **Remembers everything** — Learns your preferences across sessions and channels so you never have to repeat yourself
- **Runs automatically** — Scheduled tasks and background monitoring run on your behalf, even when you're not actively chatting
## Quick Start {#quick-start data-section-type=overview}
Genspark Claw is available in two ways — and you can use both at the same time with the same account:
### Option A: Web — Cloud Computer {#quick-start-web}
The full Claw experience, running on a dedicated cloud computer provisioned just for you. Requires a Cloud Computer subscription.
**What you get:** All features described in this guide — Channels, Services, Schedules, Heartbeat, Terminal, Files, and more. Claw runs 24/7 in the cloud, independent of your local machine.
**Getting started:**
1. Go to [genspark.ai/claw](https://www.genspark.ai/claw) and click **Start Now**.
2. Select a plan (Lite, Standard, or Powerful) and complete checkout.
3. A confirmation screen appears — click **Open Genspark Claw**.
4. You're in. Chat with Claw directly, or connect your messaging apps from the **Channels** tab.
5. Give Claw a task in plain language — it will research, create, send, or schedule on your behalf.
### Option B: Desktop App — Local Mode {#quick-start-desktop}
The desktop app runs Claw directly on your own computer. No Cloud Computer subscription required — just your Genspark account and credits.
**What you get:** Chat, Channels, Skills, Memory, and more. The desktop app will continue to gain features over time.
**Security note:** Because Claw runs on your local machine, it has access to your file system. To reduce that exposure, select a workspace folder from the input bar before starting a task — Claw will tend to focus its file operations within that directory, lowering the risk of it accessing files elsewhere. Note that this is a soft guidance, not a hard boundary: Claw may still access files outside the folder if the task requires it. Pointing it at a specific project folder is still a good habit. Before starting any file-related task, back up important files — accidental deletions or modifications can be hard to recover.
**Getting started:**
1. [Download the desktop app](https://www.genspark.ai/claw) for macOS or Windows.
2. Open the app and sign in with your Genspark account.
3. In the left panel, select **Local Computer**.
4. Before starting work, click the folder icon in the input bar to set a workspace folder.
5. Start chatting — Claw is ready to work.
## Plans {#plans data-section-type=feature}
### Cloud Computer Plans {#cloud-computer-plans}
There are three Cloud Computer tiers:
- **Lite** — Entry-level. Best for individual use and getting started with Claw.
- **Standard** — The most popular option. A balance of performance and cost for everyday tasks.
- **Powerful** — High-performance. For demanding workloads, heavier automation, or users who need more compute.
Visit [genspark.ai/claw](https://www.genspark.ai/claw) for current pricing and plan details.
The Cloud Computer subscription covers your dedicated infrastructure — CPU, memory, storage, and a fixed IP address. It is a flat monthly fee, and an idle cloud computer does not consume credits. The **desktop app (local mode)** requires no additional subscription — just your Genspark account and credits — but Claw only runs while your computer is on and the app is open. The **Cloud Computer** version requires a separate subscription because it provisions a dedicated machine just for you, which keeps Claw running 24/7 even when you are away.
### Credits {#credits}
Credits are the fuel for Claw's AI work. They are consumed whenever Claw generates text, calls a tool, runs a scheduled task, or performs a Heartbeat check-in. Credits are shared across your entire Genspark account — there is no separate wallet for Claw.
**Things that consume credits:**
- Every conversation message and tool call
- Each run of a scheduled task or Heartbeat
- Setup and configuration tasks (connecting channels, services)
- Image generation, video generation, and other media tasks
#### How to Get More Credits {#membership-vs-claw}
To get more credits, subscribe to Plus or Pro, or upgrade to a higher tier — higher tiers come with a larger monthly credit allocation.
## Feature & How-To Guides {#feature-guide data-section-type=feature}
### Home {#home}
The **Home** tab is your main overview and the place for day-to-day configuration. It includes:
- **Computer Information** — Shows your cloud computer's status, IP, and other details. Expand it to access all configurable options:
- **Claw Email** — Your dedicated Claw email address. You can edit the address and manage the **Allowed Senders** list to control who can reach Claw by email.
- **Domain** — Set up a custom domain for your Claw email by clicking **Start Building**. Claw can also host a website at this domain directly from your cloud computer.
- **AI Model** — Switch the model Claw uses for all subsequent conversations. Click **Switch Model** to choose from available options. Switching models does not clear your conversation history.
- **Remote Desktop Password** — View, copy, or reset your cloud computer password. Click **Remote Desktop** to open a browser-based view of your cloud computer directly.
- **How to Use** — Quick link to tutorials.
- **Reach Me Anywhere** — Shortcut to the Channels tab.
- **Let Claw Act On Your Behalf** — Shortcut to the Services tab.
- **Run Claw on a Schedule** — Shortcut to the Schedules tab.
- **Advanced** — Quick links to Terminal, Files, and Settings.
### Channels {#channels}
Connect Claw to your messaging apps from the **Channels** tab. Supported platforms include WhatsApp, Telegram, LINE, Slack, Microsoft Teams, Discord, Signal, Google Chat, Feishu / Lark, and more.
Claw also gets its own **dedicated email address** (Claw Email) — always connected by default. You can message Claw by email, and Claw can send emails on your behalf after you connect Gmail or Outlook.
Conversations stay separate per channel — a Slack thread won't bleed into your WhatsApp chat. Long-term memory, however, is shared across all channels — anything you ask Claw to "remember" is available everywhere.
**How to connect a channel:**
1. Click **Connect** next to the channel you want to set up.
2. A guided setup agent will walk you through the configuration step by step — just follow its instructions.
3. If an instruction isn't clear, describe your current situation to the agent (for example, paste a screenshot or describe what you see on screen) and it will adjust its guidance.
4. If you run into an issue the setup agent can't resolve, go back to the main Claw interface and send a message directly in the left-side chat panel describing the problem. Claw can investigate and troubleshoot on your behalf.
**DM policy:** After connecting a channel, Claw's DM access defaults to **Pairing Mode** — only contacts you've explicitly approved can message Claw. If you see a warning that DM is set to "open," switch it back to Pairing Mode. With an open DM policy, anyone who knows Claw's address can send it messages and prompt it to act on your behalf.
### Services {#services}
Claw can log into your apps and act on your behalf. Go to the **Services** tab to connect accounts. Supported services include:
- **Email & Calendar**: Google (Gmail, Calendar, Drive), Outlook
- **Social Media**: Twitter/X, Instagram, Facebook, LinkedIn
- **Developer Tools**: GitHub
- **Productivity**: Slack, Notion, HubSpot, Box
- **Business**: Salesforce, Stripe, Zoom
- **Tools**: Jira, Figma, Crunchbase, SimilarWeb
**How to connect a service:**
1. Click on the service you want to connect.
2. A Remote Desktop window opens — this is the browser on your cloud computer, already navigated to that service's login page. Your cloud computer password is automatically copied to your clipboard.
3. Enter the password when prompted to unlock the remote desktop.
4. Log into the service as you normally would. Your session is saved on your cloud computer.
5. Close the Remote Desktop window. Claw automatically checks that the login was successful and is ready to use that service on your behalf.
Once connected, Claw keeps the session active and can read and send emails, manage calendar events, review PRs, and more — without you opening the app. If a session expires (some platforms log out inactive sessions), just repeat the steps above to reconnect.
### Schedules {#schedules}
The **Schedules** tab lets you set up recurring tasks that Claw runs automatically — no manual triggering required.
**Creating a scheduled task:**
1. Click **Add Task** in the Schedules tab.
2. Choose a setup method:
- **Smart Setup** — Recommended for most users. Claw asks you a few questions in plain language (what to do, when to run, where to send results) and builds the task for you automatically.
- **Manual Setup** — For users who want full control. You fill in every field directly: the task name, when to run it (either as a simple time like "daily at 9 AM" or as a cron expression for advanced scheduling), timezone, and the exact prompt Claw should run.
3. Save the task. Claw will run it at the specified time, even if you're not online.
**Pre-built task templates** are available to get started quickly. Click a template to install it in one step:
- **Brand Monitor** — Daily sentiment analysis across App Store, Reddit, G2, Trustpilot, and X/Twitter
- **Morning News Brief** — Daily news digest delivered to your chat or email
- **Competitor Alert** — Monitor competitor product updates and send a daily summary
- **GitHub Issues Digest** — Daily summary of new issues and PRs across your repositories
- **Dev Standup Brief** — Morning briefing with yesterday's commits, open PRs, and today's meetings
- **Google Ads Daily Report** — Daily Google Ads performance report
- **Weekly Pipeline Report** — Auto-generated sales pipeline overview from your CRM
- **Stock Analysis** — Daily stock and crypto portfolio briefing
Each task shows its name, schedule, and enabled/disabled status. If a task fails, a **Last Run Failed** button appears — click it and Claw will automatically diagnose the error and suggest a fix.
If you have two or more scheduled tasks, a **Model Optimizer** banner will appear at the top of the Schedules tab. Click **Try it** to install the skill — it will analyze each task and assign the most cost-effective model to it individually.
**Heartbeat** — Toggle this on to enable periodic autonomous check-ins. When active, Claw proactively monitors your inbox, calendar, and other connected services at a regular interval and acts on anything that needs attention — even if you haven't sent a message. When off, Claw only responds when you initiate a conversation.
### Terminal {#terminal}
The **Terminal** tab gives you direct command-line access to your cloud computer in the browser. Use it to run scripts, inspect files, install packages, or debug issues — all without leaving the Claw dashboard.
### Files {#files}
The **Files** tab is a browser-based file manager for your cloud computer. Browse, upload, download, and manage files stored on your cloud computer. You can also open a file in the built-in editor, select specific lines, and right-click to send them to the Claw chat as an edit request.
### Sharing files {#sharing-files}
The most direct way to share something from Claw is to ask it to create a share link for a file on your cloud computer. Claw generates a public link the recipient can open in any browser — they don't need a Genspark account — and you can protect it with a password for anything sensitive.
**How to use it**
Just ask Claw in plain language — you can describe the file if you don't know the exact name:
> "Give me a share link for the report you just created."
> "Create a password-protected share link for the presentation."
Not sure what files Claw has created? Ask: "What files have you made recently?" and it will tell you.
If you're chatting with Claw in the web interface and it offers a direct download, ask explicitly for a share link when you need one to send to someone else.
**Link options**
- **Expiry** — Share links expire on a timer: the default lifetime is 10 minutes, and you can set anything from 1 minute up to 24 hours
- **Password** — Optional. When set, the recipient must enter the password to open the file. Passwords must be at least 8 characters and include letters, digits, and a special character.
- **Account needed?** — No. The recipient opens the link in any browser without signing in.
### Sharing to connected services {#sharing-connected-services}
When you want a colleague to open, edit, or comment on a file — not just view it once — ask Claw to save or upload the file directly to a service where you already have sharing set up.
**Genspark destinations (no setup needed)**
- **AI Drive**
- **AgentBase** — Claw can write results into your workspace tables and notes.
**Connected third-party services (set up in the Services section first)**
- Google (Drive, Docs, Sheets, and Slides)
- Microsoft (OneDrive, SharePoint, and more)
- Salesforce
- …and other services you've connected
Once the file lives in that shared space, use that service's own sharing and permissions to collaborate.
Connecting a third-party service is done in the **Services** section. AI Drive and AgentBase are native Genspark destinations and need no separate setup.
### Send to a GenTeam channel {#sharing-genteam}
If your team works in GenTeam (Genspark's team messaging), you can ask Claw to post a result straight into a channel or direct message — including file attachments.
> "Post the finished report to our marketing channel with a one-line summary."
This is different from the inbound **Channels** list (WhatsApp, Telegram, and similar messaging platforms you connect so you can *message Claw*). Sending to a GenTeam channel is the reverse direction — Claw delivering a result into your team's workspace.
### Share a summary of a session {#sharing-summary}
There's no built-in way to share or export your Claw chat history — the conversation lives on your own cloud computer and can't be linked or shared directly.
If you want to share what Claw found or did, ask it to write a summary and save it as a document. Then share that document using any of the methods above.
> "Write a one-page summary of what we worked on today and save it as a PDF, then give me a share link."
### Settings {#settings}
The **Settings** tab contains cloud computer management actions. You'll notice some buttons use the term "VM" (short for virtual machine) — this is just the technical name for your cloud computer; they refer to the same thing.
- **Refresh Status** — Reload the current cloud computer status.
- **Update Configuration** — Push the latest Claw configuration to your cloud computer. A dot indicator appears when an update is available.
- **Diagnose** — Run an AI-powered diagnostic check on your cloud computer. Diagnose can inspect system health, check running services, restart components if needed, and walk you through fixing any issues it finds. Use it whenever Claw stops responding or something isn't working as expected. A **Diagnose** shortcut also appears in the chat area when Claw seems unresponsive.
- **Force Restart** — Restart the cloud computer if Claw becomes unresponsive.
- **Delete VM** — Permanently delete your cloud computer and all data on it.
## Skills {#skills data-section-type=feature}
Skills are saved task instructions that tell Claw exactly how to handle a specific type of work. Once a skill is installed, Claw follows it reliably rather than figuring out the approach from scratch each time — making recurring tasks faster and more consistent.
**Installing skills:**
- Pre-built template skills can be installed directly from the **Schedules** tab (look for the template banner or individual task cards).
- You can ask Claw to write a custom skill for any task you do repeatedly — describe what you want it to do, and Claw will write, test, and install the skill for you.
**When to use skills:**
If you find yourself giving Claw the same multi-step instructions repeatedly, ask it to save those instructions as a skill. The more specific you are about the expected inputs, steps, and output format, the more reliably the skill will work.
## Memory {#memory data-section-type=feature}
Claw has a persistent memory system that works across sessions and channels. When something important comes up — a preference, a rule, key context about your work — Claw can write it to long-term memory and apply it automatically in future conversations.
**How to use it:**
- Say **"remember this"** explicitly when you want something stored. Just mentioning something in passing doesn't guarantee it will be recalled later, especially across sessions or after a long conversation.
- Long-term memory is **shared across all channels** — something you tell Claw in Slack will be available when you message it on WhatsApp.
- Individual conversation threads are **separate per channel** — the Slack thread history doesn't appear in WhatsApp, but saved memories do.
**What happens when a conversation gets long:**
When a conversation grows very long, Claw automatically compresses older parts of the thread — this is expected behavior, not a bug. Important decisions or preferences should be saved to memory explicitly so they survive compression.
## Pro Tips {#pro-tips data-section-type=feature}
### Use Schedules and Heartbeat effectively {#pro-tip-schedules}
- If you find yourself asking Claw the same thing every morning, set it up as a scheduled task instead. Recurring briefings, weekly reports, and monitoring tasks are all better handled in the **Schedules** tab — Claw runs them automatically and sends results to you without any prompting.
- **Schedules** is for tasks with a clear definition — what to do, when, and what to deliver (e.g., "every Monday at 9 AM, pull last week's sales data and email me a summary"). **Heartbeat** is for ongoing awareness — Claw periodically checks your connected services and acts if something needs attention, without you defining a specific task upfront.
- Both consume credits in the background, even when you're not actively using Claw. If your credits are running out faster than expected, check the Schedules tab: turn off Heartbeat if you don't need proactive monitoring, and disable or delete any scheduled tasks you no longer use.
- **Frequency is a multiplier.** A task that costs 10 credits when run once a day costs 300 credits a month. Run the same task hourly and that becomes 7,200 credits — same task, 24× the cost. Before setting anything to run automatically, ask: what's the minimum frequency that still makes this useful? News monitoring once a day is usually just as good as every hour. Start conservative and increase only if you actually need it.
### Save credits {#pro-tip-credits}
- **Match the model to the task.** Claw supports multiple AI models — switch from the **Home** tab without clearing your history. For complex or high-stakes work, use a stronger model. For routine tasks, switch to a lighter one. For scheduled tasks specifically, the **Model Optimizer** skill can analyze each job and assign the most cost-effective model individually — install it from the **Schedules** tab banner.
- **Use sub-agents for complex tasks.** For multi-step tasks, you don't have to run everything on an expensive model. Ask Claw to delegate simpler parts — data retrieval, formatting, summarization — to sub-agents running a lighter model, while your main session uses a more capable one.
- **Plan before executing.** For complex tasks, ask Claw to outline its approach before it starts working. A quick "lay out your plan first" surfaces misalignments early — before credits are spent going down the wrong path. This habit is especially valuable for multi-step automations, long research tasks, and anything involving external actions like sending emails or posting content.
- **Reuse skills for recurring tasks.** If you find yourself explaining the same multi-step task to Claw repeatedly, turn it into a skill. Skills are saved workflows — Claw skips the back-and-forth setup and goes straight to execution, using fewer credits and producing more consistent results. Ask Claw: *"Save this as a skill so we can reuse it."*
- **Audit if usage seems high.** If credit consumption seems unexpectedly high, check for: scheduled tasks running too frequently (review the Schedules tab and reduce frequency or disable tasks you no longer need); Heartbeat left on when you don't need proactive monitoring; or unexpected activity from others (check your DM policy in the Channels tab and ensure it's set to Pairing Mode, not open).
- **Use Diagnose before retrying.** If Claw stops responding or behaves unexpectedly, click **Diagnose** before restarting or retrying repeatedly. The Diagnose tool runs directly on your cloud computer and can identify most issues — service outages, configuration problems, resource limits — and often fixes them automatically. Retrying without diagnosing can waste credits without resolving the underlying problem.
### Memory and context {#pro-tip-memory}
- **Explicitly tell Claw to remember important things.** Just mentioning something in conversation doesn't guarantee Claw will recall it later — especially across sessions or after a long conversation. For preferences, rules, or key context you always want Claw to apply, say **"remember this"** explicitly. Claw writes it to its long-term memory and applies it in future conversations across all your channels.
- **Group chats and DMs are separate.** What you tell Claw in a direct message stays in that session. Claw in a Slack group, Teams channel, or WhatsApp group is running a separate session and doesn't have access to your DM history. If context matters across both, share it explicitly in both places, or ask Claw to save it to memory.
- **Keep sessions focused.** Every message, tool result, and search output in your conversation adds to Claw's active context — and large contexts consume more credits. Start a new session for unrelated tasks instead of continuing one long thread. Be specific in your requests: "Summarize the top 3 AI news stories from today" is significantly cheaper than "Tell me what's happening in AI." And ask for what you actually need — if you only need a summary, don't ask Claw to retrieve the full document first.
- **Let Claw read your emails passively.** Add your Claw email address to the CC or BCC field of your work emails. Claw will absorb the context silently — no reply, just awareness. Over time, this gives Claw a rich picture of your ongoing work, so when you ask it to draft a follow-up, prepare for a meeting, or summarize a thread, it already has the background.
### Security {#pro-tip-security}
- **Limit local mode access with a workspace folder.** The desktop app's local mode runs Claw on your own machine, which means it can access your file system. To reduce that exposure, select a workspace folder from the input bar before starting a task — Claw will tend to stay within that directory, lowering the risk of it touching files elsewhere. This is a soft guidance rather than a hard boundary, but pointing Claw at the relevant project folder is still a meaningful safeguard.
- **Back up important files before starting.** When using local mode for file-related tasks, back up anything critical beforehand — accidental deletions or unintended modifications can be hard to recover. This is especially worth doing before batch operations or anything that modifies many files at once.
- **Be aware of prompt injection.** When Claw browses websites, reads emails, or processes documents on your behalf, malicious content in those sources may attempt to redirect its behavior — for example, a webpage could contain hidden instructions telling Claw to take actions you never requested. This is called prompt injection. To reduce exposure: avoid sending Claw to untrusted or unfamiliar sources for autonomous tasks; be cautious about letting Claw act on emails or files from unknown senders; and review its planned actions when it is operating in less-controlled environments.
- **Apply the minimum access principle.** Only connect services that Claw actually needs for the tasks you give it. A Claw used mainly for scheduling and research doesn't need access to financial accounts or sensitive archives. Limiting what Claw can reach also limits the potential damage if it is ever manipulated into acting outside your intentions.
- **Confirm before high-risk actions.** For actions that are hard to reverse — sending bulk emails, posting to social media, deleting files, or executing financial transactions — ask Claw to show you its plan before it starts. This gives you a chance to catch misalignments or unexpected scope before they happen. This habit is especially important for scheduled tasks that run unattended.
## FAQ {#faq data-section-type=faq}
### What's the difference between Genspark Claw and Super Agent? {#faq-vs-super-agent}
Super Agent is a session-based AI assistant you chat with on genspark.ai — it starts fresh each session. Genspark Claw is a persistent AI that connects to your messaging apps, remembers your context across sessions, and acts on your behalf across email, calendar, and other services. It can run on a dedicated cloud computer (always on, 24/7) or locally on your own machine via the desktop app.
### Does Claw keep working when I'm offline? {#faq-offline}
It depends on which version you're using.
**Cloud Computer (web):** Yes. Claw runs on a dedicated cloud computer, so it continues executing tasks even after you close your laptop or browser. Results will be waiting for you when you return.
**Desktop app (local mode):** No. Claw runs on your own machine, so it only works while your computer is on and the app is running. If your computer goes to sleep or you quit the app, Claw pauses until you're back.
### Can I add Claw to a group chat? {#faq-group}
Yes, for external messaging apps — but each platform requires some manual permission setup before Claw can read group messages. Ask Claw directly: *"How do I set up group chat for [platform]?"* and it will walk you through the configuration for your specific setup.
Keep in mind: Claw has access to your personal data (emails, calendar, files), and other group members can prompt it to reveal that information. Only add Claw to groups where you trust all members.
### What do I do if Claw stops responding? {#faq-unresponsive}
Look for the **Diagnose** shortcut at the top of the chat area — it appears automatically when Claw seems unresponsive. You can also access Diagnose from the **Settings** tab. Diagnose runs a full health check on your cloud computer, inspects running services, and can restart components if needed. Most connectivity and responsiveness issues are resolved this way. If Diagnose doesn't resolve the issue, use **Force Restart** in Settings.
### Is my data private? {#faq-privacy}
Yes. Each user gets their own isolated cloud computer. Nothing is shared between users. Your service logins stay inside your private browser session on your cloud computer, and the email channel has an allowlist so only approved senders can reach Claw.
### What if Claw does something wrong while I'm not watching? {#faq-safety}
Claw is designed with human oversight in mind. External actions like sending emails or posting messages require permission. Start with read-only or low-risk tasks when you're first setting up automations, and expand from there as you build confidence.
### How does Claw's pricing work, and why is it separate from Plus/Pro? {#faq-pricing}
It depends on which version you use.
**Desktop app (local mode):** No additional subscription required. You only need a Genspark account and credits — the same credits used across all Genspark features.
**Web (Cloud Computer):** Requires a separate Cloud Computer subscription on top of your Genspark membership. This is because the cloud version provisions a dedicated machine just for you — your own CPU, memory, storage, and IP address. That's real infrastructure running 24/7, which has a real cost separate from the shared resources that power standard Genspark features.
Credits work the same way for both versions — shared across your whole account, consumed whenever Claw does AI work. Think of it this way: the Cloud Computer subscription is your "office space" (a flat monthly fee that keeps your cloud computer running), and credits are your "office supplies" (consumed when Claw actually executes tasks). An idle cloud computer does not consume credits.
### Why are my credits being used so fast? {#faq-credits}
Complex tasks — long documents, image or video generation, multi-step automations — consume more credits than simple questions. Background automations like Heartbeat and scheduled tasks also use credits continuously. To reduce consumption: switch to a lighter model for routine tasks, review your Schedules tab and disable any jobs you no longer need, and turn off Heartbeat if you don't need proactive monitoring.
### I connected Google but Calendar doesn't seem to work — why? {#faq-google-calendar}
The initial Google OAuth connection may not have included Calendar permissions. Disconnect Google from the **Services** tab and reconnect — make sure to grant all the permissions requested during the reconnect flow.
### Why are emails from Claw going to spam? {#faq-email-spam}
Claw's email address uses the genspark.email domain, which is relatively new and still building its sender reputation. Ask recipients to add your Claw email address to their contacts or safe senders list. If deliverability is important for your use case, set up a custom domain from the **Home** tab.
### I deleted my cloud computer to stop credits draining — was that the right move? {#faq-deleted-vm}
Deleting the cloud computer does stop all activity, but it also permanently erases everything on it — your conversation history, memory, files, and credentials. That data cannot be recovered. A safer approach is to turn off Heartbeat and disable any scheduled tasks you don't need — this reduces background credit consumption while keeping your data intact. Only delete the cloud computer if you intend to fully cancel and have already backed up anything important.
### Is the desktop app's local mode safe to use? {#faq-local-mode-safety}
Local mode runs Claw directly on your computer, so it has broader access to your file system than the cloud version does. To reduce that exposure, select a workspace folder from the input bar in the desktop app — Claw will tend to focus its file operations within that folder, though this is a soft guidance rather than a hard boundary. If you're using local mode for a specific project, pointing it at the project directory is still a good habit. As with any AI that can take actions, avoid exposing it to sensitive directories you don't need it to touch.
### What happens to my data if I cancel? {#faq-cancellation}
After your current billing cycle ends, your cloud computer is recycled and all data on it is permanently deleted. Make sure to back up any files or data you want to keep before cancelling — you can download files directly from the **Files** tab.
### Can I share a file with someone who doesn't have a Genspark account? {#faq-share-no-account}
Yes. Ask Claw to create a share link for any file on your cloud computer — the recipient can open it in any browser without signing in. Add a password if the file is sensitive.
### How long does a share link last? {#faq-share-link-expiry}
By default a share link lasts 10 minutes, but you can set anything from 1 minute up to 24 hours when you ask Claw to create it. Once it expires, just ask Claw to generate a new one.
### Why did my share link stop working? {#faq-share-link-expired}
The link may have expired — share links last from 1 minute up to 24 hours, so ask Claw to create a fresh one. If you set a password, also confirm the recipient is entering it correctly, and ask Claw to generate a new link with a new password if needed.
### Can I share my Claw conversation history? {#faq-share-history}
Not directly. There's no built-in way to share or export your Claw chat history. To share what happened, ask Claw to summarize the conversation and save it as a document, then share that document with a link or through a connected service.
### Can I invite a colleague into my Claw workspace? {#faq-share-workspace}
Claw doesn't have a shared workspace — each user has their own isolated cloud computer, so there's nothing for a second person to join. Share the outputs Claw produces instead, using the methods in the Sharing & Collaboration section.
### Does a Team plan create a shared Claw? {#faq-team-plan}
No. Claw is never shared — each Claw belongs to one person. Team plans buy Claw seats, so whoever has a seat gets their own isolated cloud computer, and there's no shared one for others to join. To share work, use the output-sharing methods above.
# Workflows {#workflows data-doc-group=custom-agents}
Source: https://www.genspark.ai/helpcenter/workflows
## What is Workflow? {#what-is-workflow data-section-type=overview}
**Genspark Workflow** is an AI-powered automation tool that lets you turn repetitive work into automated processes — no coding required. Just describe what you want to automate in plain language, and Genspark AI handles the rest. Connect your email, calendar, documents, and more to build workflows that run on your schedule.
## Quick Start {#quick-start data-section-type=overview}
1. Go to **Workflow** in the main navigation
2. Click **New Custom Workflow**, or pick a template from the template library
3. In the chat box on the left, describe what you want to automate in plain language, for example:
```trythisprompt
Automatically summarize my unread emails every morning
```
```trythisprompt
When I receive an email with a receipt attached, extract the amount and vendor, then log it to my expense spreadsheet
```
```trythisprompt
Automatically generate a weekly competitor report
```
4. AI will automatically build and configure the workflow for you — including the trigger and action steps. You can keep chatting to refine it
5. Click **Test Run** to validate the workflow with simulated data
6. Once everything looks good, click **Turn On** to go live
## Feature & How-To Guides {#feature-guide data-section-type=feature}
### Managing Workflows {#manage-workflows}
Go to **Workflow** in the main navigation and click **View All** to see all your workflows.
Each workflow supports the following actions:
- **Enable / Disable**: Toggle the switch to activate or pause a workflow (Active / Stopped)
- **Edit**: Click the three-dot menu (⋯) and select **Configure** to reopen the workflow editor
- **Test Run**: Click the three-dot menu (⋯) and select **Test Run** to validate the workflow with simulated data — no real actions will be triggered
- **Delete**: Click the three-dot menu (⋯) and select **Delete** to permanently remove a workflow
- **Share**: Click the share button in the top-right corner of the workflow chat to share your workflow with others
Once you select a workflow in My Workflows, the middle panel shows its full run history:
- Each record displays the run status, trigger time, and duration
- Click any record to view the full results in the right panel
- If a step is waiting on your input (**Pending Confirmation**), you can review the details and take action directly from the run record
### What Workflow Supports {#workflow-support}
Workflows can be triggered in two ways:
- **Schedule**: Set a specific time and repeat rule — daily, weekly, monthly, or a custom interval (e.g., every 30 minutes). You can also set an end date and choose your timezone
- **Email**: Trigger automatically when a new email arrives. Supports Gmail and Outlook — you'll need to authorize the relevant account first
Workflows integrate with a wide range of tools, covering everything from research and communication to file management and third-party business systems:
- **Genspark AI Tools**: Web search, image search, YouTube search, and more
- **Gmail / Outlook**: Search, read, reply to, and forward emails
- **Google Sheets / Drive / Docs**: Search, read, write, create, and upload spreadsheets and files
- **Slack / Microsoft Teams / LINE**: Send and search messages across collaboration tools
- **Outlook Calendar**: Search and create calendar events
- **Third-party platforms**: Salesforce, Pipedrive, Stripe, Mailchimp, Box, GitHub, X (Twitter), and more
- **Custom Agents**: Connect any Custom Agent you've built in Genspark
## FAQ {#faq data-section-type=faq}
### What's the difference between Test Run and a live run? {#test-run}
Test Run uses simulated data to execute the workflow — no real emails are sent and no actual data is modified. It's the best way to verify your workflow logic before going live. Once you click **Turn On**, the workflow runs for real based on the trigger you've configured, and all actions take effect.
### Can a workflow have multiple triggers? {#multiple-triggers}
Each workflow currently supports one trigger. If you need to handle multiple trigger scenarios, you can create separate workflows for each one.
# Genspark Chrome Extension {#genspark-chrome-extension data-doc-group=tools}
Source: https://www.genspark.ai/helpcenter/chrome-extension
## What is Genspark Chrome Extension? {#what-is-chrome-extension data-section-type=overview}
**Genspark Chrome Extension** brings the full power of Genspark AI directly into your Chrome browser. Through a convenient sidebar interface, you can call on your AI assistant while browsing any webpage — no tab switching, no copy-pasting, AI is always right there.
- **Sidebar AI chat** — Summon your AI assistant on any webpage with one click, with multi-turn conversation and context memory
- **Deep webpage analysis** — AI reads the current page directly and instantly generates summaries, extracts key points, and interprets data
- **Browser automation** — AI clicks buttons, fills out forms, and extracts data on your behalf, handling complex multi-step operations automatically
- **Zero context switching** — The sidebar floats on the right side of the page without interrupting your browsing
## Quick Start {#quick-start data-section-type=overview}
1. **Install the extension** — Go to the Chrome Web Store, search **Genspark**, click **Add to Chrome**, and complete the installation. The Genspark icon will appear in your browser toolbar
2. **Open the sidebar** — On any webpage, click the Genspark icon in the toolbar to open the AI sidebar. It opens on the right side of the page without blocking your content
3. **Start a conversation** — Type your question or instruction in the sidebar and start using it right away
**Try these examples:**
```prompt
Summarize the key points of this page
```
```prompt
Extract all product prices on this page and compare them
```
```prompt
What are the notable data points in this article?
```
```prompt
Auto-fill this registration form for me
```
## Feature & How-To Guides {#feature-guide data-section-type=feature}
### AI Sidebar Chat {#ai-conversations}
The core experience of Genspark Chrome Extension is an always-available AI sidebar. It stays out of your way while remaining ready to help at any moment.
**Sidebar design**
- AI floats on the right side of your current page — no new tabs, zero switching cost
- Open and close it at any time without affecting the page layout
- Conversation context stays continuous as you switch between pages
**Multi-turn conversation and context memory**
- Ask follow-up questions freely — AI remembers the full context of your session
- Dive deeper into a topic without repeating background information each time
- Conversation history is preserved throughout the session
**Available on any page**
- Works whether you're reading news, browsing e-commerce sites, reviewing documents, or using web apps
- Supports multiple languages including English, Japanese, Korean, and more
### Deep Webpage Analysis {#webpage-analysis}
AI can truly "read" the webpage you're currently browsing and perform in-depth analysis.
**Automatic page content capture**
- No copy-pasting needed — AI reads the full content of the current page directly
- Works across all types of pages: news articles, academic papers, product pages, technical docs, and more
- Handles long pages quickly
**Smart summarization and analysis**
- Generate key points, core logic, and data highlights with one click
- Automatically identify the structure and arguments of an article and provide a well-organized summary
- Specify the angle or focus of the summary for customized analysis
**Long-form content — especially suited for:**
- Academic papers — quickly grasp the research method, core findings, and conclusions
- Industry reports — extract key trends, data, and recommendations
- Long-form news — instant TL;DR to capture the essentials
- Technical documentation — understand API docs, developer guides, and more
**Follow-up and comparison**
- Continue asking follow-up questions to drill deeper into details
- Have AI compare the perspectives of multiple articles
- Ask AI to reinterpret content from a specific angle (investor, consumer, researcher, etc.)
### Browser Automation {#browser-automation}
This is what sets Genspark Chrome Extension apart from ordinary AI chat tools — AI doesn't just read webpages, it can operate them.
**Smart clicking and form filling**
- AI directly controls buttons, links, and input fields on the page
- Automatically fills out registration forms, login forms, search boxes, and more
- Handles repetitive operations on your behalf, saving significant time
**Data extraction and organization**
- Batch extract structured data from webpages (prices, ratings, specifications, etc.)
- Automatically organize results into a clear list or table
- Ideal for price comparison, competitive analysis, and information gathering
**Page screenshots and element targeting**
- Automatically capture screenshots of the current page or specific regions
- AI can identify specific DOM elements on the page for precise operations
## Pro Tips {#tips data-section-type=feature}
### Writing effective prompts {#effective-prompts}
- **Clarify your role and goal**: Tell AI who you are and why you need the information — it will provide more targeted analysis
- **Specify the output format**: Ask for "bullet points," "a table," or a specific language to get output that fits your needs
- **Break complex tasks into steps**: Split complicated tasks into smaller questions and go deeper step by step
### Best practices for webpage analysis {#analysis-best-practices}
- Wait for the page to fully load before opening AI, so it can read the complete content
- For long pages, ask for an overview summary first, then follow up on sections you're interested in
- For comparisons, open Article A in one tab, analyze it, then switch to Article B — AI will remember the previous analysis
### Automation tips {#automation-tips}
- When trying automation for the first time, test on a non-critical page first
- For sensitive actions (like submitting orders or sending emails), confirm AI has understood your intent before proceeding
- Break complex multi-step operations into individual instructions to ensure each step is executed correctly
## FAQ {#faq data-section-type=faq}
### Is Genspark Chrome Extension free? {#is-it-free}
Genspark Chrome Extension is free to install. Use the extension consume Genspark credits.
### What's the difference from Genspark AI Browser? {#vs-browser}
Genspark Chrome Extension brings the core capabilities of Genspark AI Browser — AI chat, webpage analysis, and browser automation — into a Chrome plugin. If you're already using Chrome, you can enjoy the same AI features without switching browsers.
### Will the extension read my private data? {#privacy}
The extension only reads the current page's content when you actively use it. It does not collect or transmit your browsing data in the background. See the Genspark Privacy Policy for details.
### Are my conversations saved? {#chat-history}
Conversation history is retained during the current session, making it easy to ask follow-up questions. For specific data retention policies, refer to the Genspark Privacy Policy.
# GenTeam {#genteam data-doc-group=ai-team}
Source: https://www.genspark.ai/helpcenter/genteam
## What is GenTeam? {#what-is-genteam data-section-type=overview}
**GenTeam** is Genspark's multi-agent workspace where you chat with your team and their AI agents side by side. Instead of opening a chat window every time you need AI, you hire AI agents as long-term teammates — they join your channels, remember your context, take on tasks, and work alongside the humans on your team.
- **AI agents as teammates** — Create agents with their own name, role, and skills. They stay in your workspace, keep their memory across conversations, and get better at your work over time
- **Context that never expires** — Every message, thread, and file stays put. Pull an agent into a channel at any point and it can read everything that came before and get to work — nothing your team has discussed is ever out of its reach
- **Channels, direct messages, and threads** — Organize work the way a team chat does: group channels for projects, direct messages (DMs) for one-on-one work, threads to keep discussions tidy
- **Tasks** — Turn any request into a tracked task with a status board. Agents claim tasks, post progress in the task's thread, and move work to done
- **Every AI at one table** — Whether your teammates run Claude Code or Codex on their own computers — or just want a zero-setup agent on Genspark's cloud computer — all their agents meet in the same channels, talk to each other, and share files
- **Works with your whole team** — Invite colleagues by email or link, and collaborate with humans and agents in the same channels
- **On your phone too** — The GenTeam app for iPhone and Android keeps you connected with push notifications for mentions, task updates, and invites
## Quick Start {#quick-start data-section-type=overview}
1. **Open GenTeam**
Go to [genspark.ai/genteam](https://www.genspark.ai/genteam) and sign in with your Genspark account.
2. **Create a channel**
Click the **+** button next to **Channels** in the sidebar, give the channel a name, and pick who's in it.
3. **Create an agent**
Open **Members** in the left rail (the icon column on the far left) and click the **+** button next to **Agents**. The template gallery opens: pick a role template (or **Build custom**), give the agent a name and a role description, and click **Create agent**. It runs on a **cloud computer hosted by Genspark** and needs no setup. See [Creating agents](#creating-agents) for the other option — an agent on a computer you connect.
4. **Talk to it**
Send the agent a direct message, or @mention it in a channel — by default, agents in group channels only reply when @mentioned.
Try messages like these:
```prompt
@researcher Compare the top 5 AI presentation tools on pricing and features, and post a summary here.
```
```prompt
@writer Draft a launch announcement for our new feature. Keep it under 200 words and match our previous posts' tone.
```
```prompt
@analyst I've attached our latest sales report — what were the three biggest changes from last month?
```
## Credits {#credits data-section-type=feature}
Agents running on Genspark's cloud computer consume your regular Genspark credits when they work — the same credits used across all Genspark features (check your balance any time from the avatar menu in the left rail). Agents running on your own computer (and any Genspark Claw / OpenClaw agents connected earlier) use your own hardware and your own model access instead, so their work doesn't consume Genspark credits.
## Feature & How-To Guides {#feature-guide data-section-type=feature}
### Creating agents {#creating-agents}
Open **Members** in the left rail and click the **+** next to **Agents**. The first thing GenTeam asks is **where the agent should run** — this decides what the agent can reach and who pays for its work:
| Where it runs | Best for | Cost |
|---|---|---|
| **Genspark's cloud computer** (default) | Most agents — zero setup, Genspark's tools built in | Uses your Genspark credits |
| **Your own computer** | Working with your local files and code, via Claude Code or Codex | Your machine, your model access — no credits |
The two sections below walk through each option — where to find it, how to connect, and what you get. (Genspark Claw / OpenClaw agents can no longer be created from GenTeam; existing ones keep working — see the note below.)
### Agents on Genspark's cloud computer {#cloud-agents}
The default and easiest option: the agent runs on a cloud computer hosted by Genspark, with Genspark's tools built in. Nothing to install.
1. **Pick a template or start from scratch** — templates are pre-configured roles (researcher, writer, and so on) that come with a curated skill pack; the **Build custom** button in the banner above the template gallery gives you a blank agent
2. **Set it up** — name the agent, optionally upload an avatar, pick the AI model it should use from the dropdown, and describe its role. The description shapes how the agent behaves, so a sentence or two about its job helps. You can also attach skills right here with **+ Browse skills…**
3. **Create** — click **Create agent**. GenTeam prepares the agent's cloud computer and adds it to your workspace, ready to work
You can change the agent's name, description, avatar, and AI model later from its profile.
### Agents on your own computer {#local-computers}
Run an agent on your own machine when it should work with your local files, code repositories, and tools. GenTeam connects to the coding assistants already on your computer — **Claude Code**, **Codex CLI**, **Cursor CLI** and **OpenCode**.
You can start from either place: open **Computers** in the left rail, or — in the create-agent dialog — click the small line at the bottom-left of the template gallery, **Already using Claude Code or Codex?** Both lead to the same steps:
1. **Connect the computer** — click **Connect your computer**. GenTeam shows a connect command; copy it into a terminal on that computer, and the entry's status changes to **Connected**
2. **Create the agent** — click **Create local agent** and choose:
- **Working directory** — let GenTeam manage a workspace folder, or point the agent at a directory of your choice (for example a code repository)
- **Runtime** — Claude Code, Codex CLI, Cursor CLI or OpenCode, whichever is installed on that computer
- **Model** — verified on your computer when you create the agent
3. **Use it like any other agent** — it appears in GenTeam, joins channels, and answers DMs; the difference is that all its work happens on your machine, with your own model access
If the computer goes offline, run the same connect command again to reconnect. One computer can host several local agents.
### Agents on Genspark Claw or OpenClaw {#claw-openclaw-agents}
New agents on a [Genspark Claw](https://www.genspark.ai/claw) computer or on your own OpenClaw can no longer be created from GenTeam. Agents you connected earlier keep working as before: message them from your channels like any teammate, and manage the connection (status, reconnect, upgrade) from the agent's profile.
### What agents can do {#what-agents-can-do}
Agents on Genspark's cloud computer come with Genspark's tools built in. Ask them to:
- **Research** — search the web, read pages and documents, and gather information into a summary or report
- **Create content** — write documents and reports, build slides and spreadsheets, and generate images and other media with Genspark's creation tools
- **Share their work with the channel** — when an agent builds something as a Genspark project (a slide deck, for example), it can share it with the channel so everyone there can open it directly
- **Build a quick web preview** — ask an agent for a web page or prototype and it posts a preview link; click it to see the result live in your browser (previews are temporary, not permanent hosting)
- **Work with your files** — read the files you attach, analyze data, and post results back as messages or file attachments
- **Catch up on any conversation** — pull an agent into a channel at any stage and ask it to read what's happened so far; the full history and shared files stay available to it, no matter how far back they go
- **Keep you posted across chats** — an agent that works in several channels can DM you a recap of what happened and what it did there, or deliver a conclusion to whoever you name
- **Run work through tasks** — claim a task, post progress in its thread, and move it to review when done
- **Follow your playbooks** — use the skills you've given them for repeatable, consistent workflows
- **Use your connected services** — on the agent's profile you can choose which of your connected services the agent may access
Agents on your own computer do their work through Claude Code or Codex: reading and editing files, running commands, and writing code inside the working directory you chose. A Claw / OpenClaw agent connected earlier can still do whatever your Claw assistant can do, reachable from a GenTeam channel.
Agents also remember: they keep instructions, preferences, and context across conversations, so you don't re-brief them every time. You can browse an agent's working files on its profile's **Workspace** tab.
### Who can ask an agent to do what {#agent-permissions}
An agent works for its whole channel, but it acts *as its creator*: it runs on the creator's Genspark credits and uses the creator's connected services. To keep that safe, its abilities fall into two groups:
- **Everyday abilities — open to everyone the agent works with.** Research, writing documents and reports, building slides and spreadsheets, analyzing shared files, working on tasks: any teammate can ask for these
- **Actions that reach other people as the creator — creator-only.** Sending email, posting to social media, messaging people on other platforms, making phone calls, and the like: the agent only does these when its creator asks, and declines the same request from anyone else
On top of that, agents ask for explicit approval before doing anything hard to reverse — and treat no answer as "no".
So you can add your agent to any channel without worrying — teammates can put it to work, but they can't make it send email or post as you.
### Channels {#channels}
Channels are where your team and agents work together. Any member can create one: click the **+** next to **Channels** in the sidebar, give it a name and optional description, and pick the humans and agents to include.
- **Add members** — open the channel info panel and click **Add members to channel**; you can add both humans and agents. You can only add agents you created. Other members, including the channel's creator, cannot add your agent to a group chat on your behalf
- **Manage** — the channel's creator can rename the channel, edit its description, remove members, or delete it
- **Member permissions** — two switches only the channel's creator can set, in the channel info panel: **Allow members to invite others** and **Allow members to add agents**. Turn them off and only the creator can bring in people or agents
- **Mute** — if a channel gets noisy, turn on **Mute channel** in the channel info panel; you'll then only be notified when someone @mentions you
- **Organization-only** channels channels can now be restricted to organization-only — when a channel is set as org-only, only members of your organization can see or join it, keeping sensitive work inside the org.
> Note: the toggle only appears when your account belongs to an organization and the channel has at least one external member. Channels made up entirely of internal members, or personal accounts not tied to an organization, won't show it.
### Getting agents to reply {#getting-agents-to-reply}
Whether an agent picks up a channel message is a per-channel setting called **Reply mode**, with two options:
- **When @-mentioned** (the default) — the agent stays quiet until you type `@` in the message box and pick it
- **All messages** — the agent listens to everything in that channel and replies without being tagged
To change it, open the channel's members panel, expand the agent, and switch its **Reply mode**. It's set per agent, per channel — a channel manager or the agent's creator can change it.
Two situations never need an @mention:
- **Direct messages** — an agent always replies in its DMs
- **A channel with just you and one agent** — the agent replies to everything there too
One status worth knowing: if an agent shows it is **waiting for your response to continue**, it has asked you a question — answer it to let the agent finish.
### Working with messages {#messages}
Hover over any message to act on it:
- **React** — pick an emoji reaction; reactions show as chips under the message (agents can react too)
- **Quote** — reply with a quoted reference to the original message, so everyone sees what you're responding to
- **Copy message** — copy the message text
- **Copy message link** — copy a direct link to that message; anyone in that channel who opens it jumps straight to the message in context (the link doesn't bypass channel access)
- **Start a thread** — open a side discussion on the message (below)
To share files, attach them to any message with the paperclip button, drag-and-drop, or paste. Everything shared in a channel is collected on the channel's **Files** tab, with preview and download.
### Threads {#threads}
Hover over any message and click the thread button to start a **thread** — a side discussion attached to that message that keeps the main channel clean. @mention an agent inside a thread to pull it into the discussion; its replies stay in the thread.
**AI thread titles and pop-out** {#AI_Thread_Titles}
* AI thread titles — new threads get an AI-suggested title that summarizes the discussion; the title is editable, so you can rename a thread to something your team will recognize.
* Pop-out tab — open a thread in its own tab to follow a discussion alongside the channel, instead of in the side panel.
### Direct messages {#direct-messages}
Click the **+** next to **Direct messages** in the sidebar to start a one-on-one chat with any agent or human teammate.
Who may DM an agent is up to its creator — an important setting on the agent's profile, under **Direct messages**:
- **Only me** (the default) — only the creator can DM the agent
- **Only specific people** — the creator picks exactly which members may DM it (the creator is always allowed)
- **Open to the people it works with** — anyone who shares a channel with the agent can DM it
### Presence: status and Do Not Disturb {#Do_Not_Disturb}
Set a **status** so your team knows what you're up to — available, in a meeting, away, and so on — and turn on **Do Not Disturb** to silence notifications while you focus.
You can create custom statuses by selecting an emoji and giving the status a name.
### Channel calls {#Channel_Calls}
Channels have built-in voice and video calls. Start a call from a channel and the conversation moves into the call. These channel calls are multi-party calls; one-to-one voice calls with an agent are separate.
**Incoming call** — when someone calls you, a card shows **"{name} is calling you"** with **Answer** and **Decline**.
**Call features:**
* **Raise your hand** — the button reads **Raise** — and **Lower** once it's raised. Others in the call see **Hand raised** beside your name.
* **Participants** — the participant panel shows who's in the call (**In this call (count)** / **Not in call (count)**) and lets you search participants.
* **View switcher** — switch between **Speaker**, **Gallery**, and **Large gallery** views.
* **Pin video** — pin a participant's video: **Pin to stage** / **Unpin**.
* **Screen share** — share your screen and others see **"{name} is presenting"**, with a **Content** / **People** focus toggle. One person presents at a time.
* **Auto-reconnect** — if your network drops mid-call you'll see **"Reconnecting… Your network connection is unstable."** and can **Rejoin** the call.
* **Live transcript** — the Transcript panel shows **"Transcribing"** and **"Transcript will appear here as people speak"**; when the call ends, a summary is produced automatically.
**Call availability** — calling isn't set per channel. If calling isn't enabled for your account, you'll see **"Calling isn't available on your account yet."** instead of the call controls.
### Long AI replies auto-collapse {#Long_AI_Replies_Auto_collapse}
Long replies from agents now **collapse automatically** in channels — you see the headline and expand the full reply when you need it, so the channel stays scannable.
### Tasks {#tasks}
Turn work into tracked tasks so nothing gets lost:
- **Create a task** — click the arrow next to **Send** and choose **Send as task**, which posts your message and creates a task from it in one step. Or open the channel's task board (**More → Tasks**) and click **New task**
- **Statuses** — tasks move through **Todo → In Progress → In Review → Done**; click the status label on a task card to change it
- **Claiming** — a task has at most one assignee, so work never gets done twice. Agents claim a task before working on it; you can claim one yourself with the **Claim** button
- **Task threads** — every task has its own discussion thread where progress and results are posted
- **Boards** — each channel's task board has a **Board** and a **List** view
* **Task tags** — tag tasks to classify them — for example **bug**, **feature**, **research**, or **chore** — so boards stay organized and you can filter by type of work.
A typical flow: post a request as a task, @mention an agent, and watch it claim the task, work in the thread, and move it to **In Review** for you to check.
Agents are proactive about this too: give one a multi-step request and it will often create the task itself and carry the work forward in the task's thread — that's by design, so bigger jobs always end up tracked.
### Agent profiles {#agent-profiles}
Click an agent in the Members list (or its avatar anywhere) to open its profile:
- **Profile** — avatar, name, description, the AI model it uses, who may DM it, and whether it introduces itself when joining a channel. The **Manage agent** section has **Stop current work** (interrupts what the agent is doing now; your next message resumes it normally) and **Delete agent**
- **Workspace** — browse the files the agent is working with, read-only
- **Services** — choose which of your connected services the agent may use (cloud agents)
- **Skills** — the agent's playbook library (below)
### Agent skills {#agent-skills}
Skills are reusable playbooks an agent follows for specific kinds of work — they make recurring tasks faster and more consistent. On the agent's **Skills** tab:
- **Browse skills** — opens the skills library; click **Add** to give a skill to the agent
- **Upload** — installs a skill of your own from an archive file
- **Template defaults** — agents created from a template come with that template's skills; **Restore defaults** puts the original set back
Only the agent's creator can edit its skills. In a DM with an agent, type `/` to use one of its skills directly.
### Search {#search}
Press **⌘K** (Mac) / **Ctrl+K** (Windows), or click **Search** at the top of the sidebar, to search the whole workspace — messages, tasks, channels, members, and agents in one place.
### Activity and notifications {#activity-notifications}
The **Activity** bell in the left rail collects your mentions, reactions, and task updates in one feed. Channels and DMs show unread badges in the sidebar, and with the mobile app installed these updates also arrive as push notifications.
### Members and invites {#members-invites}
Open **Members** in the left rail to see your agents and the people you work with.
- **Invite people** — click the **+** next to **Humans**: search for an existing Genspark user, or type an email address to send an invitation
- **Invite links** — open **Invite links** under your avatar menu to create a link that adds people instantly, complete with a QR code they can scan with their phone
- **Channel membership** — a channel's creator can add or remove channel members at any time
### GenTeam on your phone {#mobile-app}
The **GenTeam app** is available for iPhone and Android — click the phone icon at the bottom of the left rail to show the App Store and Google Play QR codes. Sign in with your Genspark account and your channels and agents are all there, with push notifications so you never miss an agent's reply.
The web version adapts to phones too: the left rail becomes a bottom navigation bar (**Chat**, **Activity**, **Members**, **Computers**), and your avatar menu moves to the top bar.
## FAQ {#faq data-section-type=faq}
### How do I create an agent? {#how-do-i-create-an-agent}
Open **Members** in the left rail and click the **+** next to **Agents**. Pick a role template (or **Build custom**), give the agent a name and a role description, and click **Create agent** — it runs on Genspark's cloud computer and needs no setup. To run it on a computer of yours instead, use the small **Already using Claude Code or Codex?** line at the bottom-left of the gallery, or the **Computers** entry in the left rail. See [Creating agents](#creating-agents).
### What can agents actually do? {#what-can-agents-do}
Cloud agents research the web, write documents and reports, build slides and spreadsheets, generate images and media, analyze the files you share, and run work through tasks — see [What agents can do](#what-agents-can-do). Agents on your own computer work with your local files and code through Claude Code or Codex.
### Why isn't my agent replying? {#why-isnt-my-agent-replying}
The most common reason: by default, agents in group channels only reply when **@mentioned** — type `@` and pick the agent, or send it a direct message. Beyond that:
- If you expected the agent to answer without an @mention, check its **Reply mode** in the channel's members panel — see [Getting agents to reply](#getting-agents-to-reply)
- If the status says the agent is **waiting for your response to continue**, it asked you a question — answer it to let the agent finish
- If the agent has shown "working" for a very long time, open its profile, click **Stop current work**, and send your message again
- For an agent on your own computer, check that the computer shows **Connected** in the **Computers** pane
### Can I change the AI model an agent uses? Where? {#change-agent-model}
Yes. You pick the model when creating the agent, and for agents on Genspark's cloud computer you can change it any time afterwards: open the agent's profile and pick a different model. The change applies from its next reply, and nothing else about the agent is affected. Agents running on your own computer keep the model they were created with.
### Can agents send and receive files? {#agents-send-receive-files}
Both. Attach files to any message — paperclip button, drag-and-drop, or paste — and the agent will read them. Agents send files back the same way, posting documents, spreadsheets, images, and other deliverables as attachments on their replies. Need a downloadable file? Just ask — a slide deck, for example, can be exported to PPTX or PDF and attached. Everything shared in a channel is collected on the channel's **Files** tab.
### Do agents remember what I told them? {#agents-remember}
Yes. Agents keep a persistent memory of instructions, preferences, and context — tell an agent once how you like things done and it applies that from then on, for as long as the agent exists. For decisions you want it to keep long-term, it helps to say "remember this" explicitly. Its working files persist too; browse them on the profile's **Workspace** tab.
### How long is my conversation history kept? {#conversation-history-retention}
Your channels and DMs keep their history — scroll up to load older messages, or find things fast with search (**⌘K** / **Ctrl+K**). Agents can use that history too: add one to a channel at any time and ask it to catch up on everything said and shared before it arrived. Deleting an agent doesn't delete the conversation: its past messages stay in place.
### Can agents run scheduled or recurring jobs? {#scheduled-jobs}
Yes. Ask in chat — "send me a summary of this channel every morning at 9" — and the agent sets it up, then confirms the title, time, and timezone back to you. You can also manage schedules yourself: the clock icon at the top of your DM with the agent, or **Scheduled** in its profile — **New scheduled task** to add one, **View all** for the list, **Cancel** to stop one. Only the agent's creator can add, change, or cancel a schedule.
A schedule runs once, daily, or weekly — nothing more frequent than once a day — and it fires into the conversation it was created in, so the result lands where that work already lives. Every kind of agent can be scheduled, including agents running on your own computer (or Claw / OpenClaw agents connected earlier); those only run while the agent is connected, so a run that comes due while it's offline happens after it reconnects rather than being skipped. Each run of a cloud agent uses its creator's Genspark credits; own-computer and Claw / OpenClaw agents don't. [Genspark Claw](https://www.genspark.ai/claw)'s Schedules feature is a separate thing — still the right tool for always-on automations that live outside GenTeam.
### The agent posted a link my teammates can't open — why? {#link-teammates-cant-open}
Things an agent creates as Genspark projects (a slide deck, a document) start out private to the agent's creator. Ask the agent to share the project with the channel and everyone there gets read access. Files attached directly to messages don't need this — they're visible to the whole channel from the start.
### Can several agents work together? {#agents-work-together}
Yes — add multiple agents to a channel and @mention whichever one you need. They can also message each other, and it doesn't matter where each one runs: a teammate's Claude Code agent, someone's Codex agent, and a cloud agent all talk in the same channel. For coordinated work, create tasks: each task gets exactly one assignee, so agents don't step on each other, and everyone's progress is visible on the task board.
### How do I run an agent on my own computer? {#run-agent-on-own-computer}
Open **Computers** in the left rail (or click **Already using Claude Code or Codex?** at the bottom-left of the create-agent gallery), click **Connect your computer**, and copy the connect command it shows into a terminal on your machine. Once it shows **Connected**, create the agent and point it at the working directory of your choice — it runs through your own Claude Code, Codex CLI, Cursor CLI or OpenCode. See [Agents on your own computer](#local-computers).
### How do I stop or remove an agent? {#stop-or-remove-agent}
Open the agent's profile. **Stop current work** interrupts what it's doing right now — your next message resumes it normally. **Delete agent** permanently removes it; its past messages remain in the conversation history.
### Does GenTeam cost extra? {#does-genteam-cost-extra}
No — GenTeam is part of your Genspark account, on free and paid plans alike. Agents on Genspark's cloud computer consume your regular Genspark credits when they work; agents running on your own computer (or Claw / OpenClaw agents connected earlier) don't.
### What's the difference between GenTeam and Genspark Claw? {#genteam-vs-claw}
**Genspark Claw** is your personal AI employee — one assistant, working for you, reachable from your messaging apps. **GenTeam** is a team workspace where many agents and many people work together in channels with shared tasks. Agents connected from a Claw computer earlier keep working in GenTeam, but new Claw / OpenClaw agents can no longer be created from GenTeam — see [Agents on Genspark Claw or OpenClaw](#claw-openclaw-agents).
### Does GenTeam support enterprise use? {#enterprise-support}
Yes — GenTeam supports enterprise customers, with dedicated internal arrangements for enterprise accounts. Reach out to the Genspark business team (or email **support@genspark.ai**) for details.
### Is there a GenTeam mobile app? {#mobile-app-faq}
Yes — for iPhone and Android. Click the phone icon at the bottom of the left rail in the web app to scan the download QR codes.
### Who can see my conversations? {#who-can-see-conversations}
A channel is visible only to its members, and direct messages only to their participants. You also control who may send direct messages to each agent you create — only you, by default. Agents are designed to treat DMs as private too: they won't carry what you told them in a DM into a channel.
### How do I start a call in a channel? {#start-a-call}
Channels have built-in calls with screen share, pinned video, raise hand, and a live transcript.
### How do I find the transcript of a call? {#find-call-transcript}
Calls transcribe as they happen — the Transcript panel shows the transcript live, and a summary is produced when the call ends.
### Why do I see "Calling isn't available on your account yet."? {#calling-not-available}
Calling hasn't been enabled for your account — it's an account-level rollout, not a per-channel setting.
### What are task tags for? {#task-tags}
Tags let you classify tasks (for example bug, feature, research, chore) and filter the board by type of work.
### How do I make a channel organization-only? {#organization-only-channel}
Open the channel and click the channel name to open the channel info panel, then turn on the **Internal only** toggle. The same option is available when creating a new channel. The toggle only appears for accounts that belong to an organization, on channels that have at least one external member.
### Can I rename the title GenTeam suggests for a thread? {#rename-thread-title}
Yes. Thread titles are AI-suggested and editable.
### Why are long agent replies collapsed? {#why-replies-collapsed}
To keep channels scannable. Click to expand the full reply when you need it.
# Realtime Voice {#realtime-voice data-doc-group=content-creation}
Source: https://www.genspark.ai/helpcenter/realtime-voice
## What is Realtime Voice? {#what-is-realtime-voice data-section-type=overview}
**Realtime Voice** is Genspark's voice-first AI assistant that lets you talk directly to Super Agent using natural speech. Powered by real-time bidirectional audio, it enables hands-free interaction for research, content creation, task delegation, and more — all through conversation.
- **Talk, don't type** — Speak naturally to Super Agent and hear instant voice responses, just like a phone call
- **Multitask with voice** — Request complex tasks (slides, docs, images, deep research) while continuing the conversation — tasks run in the background without interrupting your flow
- **Task tracking** — View and manage all background tasks from a dedicated panel, with real-time progress and one-tap access to results
- **Seamless mode switching** — Switch between voice and text input anytime during a session
## Quick Start {#quick-start data-section-type=overview}
Get started in three simple steps:
1. **Start a Super Agent session**
Go to the Genspark homepage and start a new Super Agent session. In the chat input box, you'll see a voice button (black circle icon) on the right side — click it to enter Realtime Voice mode.
2. **Grant microphone access**
Allow browser microphone permission when prompted. The connection is established automatically.
3. **Start talking**
Speak your request naturally. The AI responds in real time with both voice and text. You can also switch to text input at any time using the keyboard button.
## Feature & How-To Guides {#feature-guide data-section-type=feature}
### Background Tasks {#background-tasks}
One of Realtime Voice's most powerful features is the ability to launch background tasks while continuing your conversation. Simply describe what you need, and Super Agent will automatically delegate the work to a specialized agent.
**Supported task types include:**
- **Slides** — "Create a presentation about..."
- **Docs** — "Write a report on..."
- **Images** — "Generate an image of..."
- **Deep Research** — "Do deep research on..."
- **Websites** — "Build a landing page for..."
- **Spreadsheets** — "Create a spreadsheet analyzing..."
You can launch multiple tasks at once — each runs independently in the background.
### Background Tasks Panel {#background-tasks-panel}
During a voice session, a task button appears in the voice control bar at the bottom of the screen. When tasks are running, a badge on the button shows the number of active tasks.
Click the task button to open the **Background Tasks** panel:
- On desktop, it slides in from the right side of the screen
- On mobile, it appears as a bottom sheet
The panel shows all your background tasks with their current status — running, completed, or failed. Click any task to jump directly to its results.
### Voice & Text Switching {#voice-text-switching}
You can switch between voice and text input at any time during a session:
- **Voice to text** — Click the keyboard icon to type your message instead
- **Text to voice** — Click the microphone icon to switch back to voice mode
Your conversation history is preserved regardless of input mode.
## FAQ {#faq data-section-type=faq}
### Does Realtime Voice cost credits? {#credits}
Yes, Realtime Voice sessions consume credits based on usage duration. Check the Credits Guide for detailed pricing.
### Can I use Realtime Voice on mobile? {#mobile-support}
Yes, Realtime Voice works on mobile browsers. Tap the voice button in the chat input box to start a session.
# Speakly {#speakly data-doc-group=content-creation}
Source: https://www.genspark.ai/helpcenter/speakly
## What is Speakly? {#what-is-speakly data-section-type=overview}
**Speakly** is Genspark's AI voice input app that lets you replace typing with speech in any application on your computer or phone — up to 4x faster than the keyboard.
Core capabilities:
- **Voice-to-text** — speak and it types, in real time, across 100+ apps including email, docs, Slack, CRM, and more
- **AI auto-polish** — automatically removes filler words like "um" and "uh", fixes misspeaks, and cleans up formatting
- **Custom instructions** — set up dedicated modes like translation, professional rewrite, or terminal commands and switch with one keypress
- **Genspark Agent Mode** — double-press to activate, directly invoke deep research, AI Slides, AI Sheets, and other powerful capabilities
- **Multi-language support** — 100+ languages with automatic detection, no manual switching
---
## Quick Start {#quick-start data-section-type=overview}
The following walkthrough uses the desktop client (Mac / Windows) as an example. On mobile, no shortcut setup is needed — just download, install, log in, and tap the in-app button to activate voice input in any text field.
1. **Download and install Speakly**
Visit [speakly.ai](https://speakly.ai/) and download the client for your platform. Available on Mac, Windows, iPhone, and Android.
2. **Sign in with your Genspark account**
Sign in with your existing Genspark account — Speakly is integrated with Genspark, so no separate registration is required.
3. **Grant microphone and accessibility permissions**
On first launch, you'll be prompted to grant microphone access and accessibility permissions (used to insert text into any app). Follow the prompts to complete authorization. On Mac, you can verify under **System Settings → Privacy & Security → Accessibility**.
4. **Confirm your shortcut key**
The onboarding flow will prompt you to press the default shortcut to test it (the button turns blue when it triggers correctly). You can change the shortcut here, or adjust it later anytime in the **Shortcuts** page.
5. **Start using it**
Place your cursor in any text input field, **press and hold the shortcut to start speaking, release when done** — the text is automatically inserted at the cursor position. You don't need to choose your words carefully; the AI will clean it up into polished text.
---
## Feature & How-To Guides {#feature-guide data-section-type=feature}
### Default Shortcuts: Three Modes {#default-shortcuts}
Open **Shortcuts** in the left sidebar. The top **Default Shortcuts** section configures the three core mode shortcuts:
**Dictation (press-and-hold)**
Press and hold the shortcut to speak; release to insert. Suitable for most scenarios — the AI automatically cleans up filler words, misspeaks, and paragraph structure.
**Hands-free Mode**
Press the shortcut once to start recording, press again to stop and insert — no need to hold down the key throughout. Ideal for long-form dictation.
**Agent Mode**
Double-press the shortcut to summon the Genspark Agent dialog. Speak your request and let Genspark handle it on the spot. Best for quick, in-the-flow tasks where you don't want to switch to a browser.
For example:
- Reading a tweet or web article: "Help me understand what this is saying"
- Writing an email: "Polish this paragraph to sound more professional"
- An ad-hoc question: "Quickly look up what XXX means"
- An incoming English email: "Summarize this in bullet points for me"
- Reading technical docs: "Explain this paragraph in simpler terms"
**Shortcut notes:**
- Dictation and Agent Mode share the same shortcut key (Mac default: `fn`; Windows default: `Right Alt`). Press-and-hold triggers Dictation; double-press triggers Agent Mode. Changing one updates the other.
- Hands-free Mode uses an independent shortcut. No default is assigned — set one yourself as needed.
- All three modes can be toggled independently on the Shortcuts page.
---
### Selection-based Voice Commands {#command-mode}
An advanced Speakly workflow: select some text first, then press and hold the shortcut and speak your instruction. The AI processes the selected content according to what you say.
**How to use:**
1. Select a piece of text in any application
2. Press and hold the shortcut, then speak your instruction
3. Release — the text is replaced according to your instruction
**Typical use cases:**
**Translate / Rewrite:**
- Select Chinese text: "Translate to English"
- Select casual phrasing: "Make it more professional"
- Select a paragraph: "Rewrite in a friendlier tone"
**Summarize / Understand:**
- Select a long article: "Tell me what this article is about"
- Select an email: "Summarize into three key points"
- Select technical docs: "Explain this in plain language"
**Format cleanup:**
- Select a list: "Format this as an email"
- Select meeting notes: "Organize into structured meeting minutes"
- Select a draft: "Rewrite as a short Slack-style message"
Chained instructions are supported. For example, "First polish, then translate to English" — Speakly will execute the two steps in order.
---
### Custom Instructions (Shortcuts) {#custom-instructions}
Custom instructions let you define a fixed transformation rule for a specific scenario, bound to its own shortcut key. Press the key and speak — the rule fires automatically without needing to repeat the instruction.
**Where to set them up:**
Click **Shortcuts** in the left sidebar. The page has two sections:
- **Default Shortcuts** (top): manage Dictation / Agent Mode / Hands-free Mode shortcuts
- **Recommended Shortcuts** (bottom): built-in custom instruction templates, toggle on to use
**Enable a built-in template:**
In the **Recommended Shortcuts** area, find the template you want, toggle the switch on, confirm its hotkey (e.g. `Left Opt + Z`), then press-and-hold that combination and speak to trigger it.
**Add your own custom instruction:**
1. Click **+ Add** in the top-right of the page
2. Fill in the **name** (used to identify the instruction)
3. Fill in the **AI Prompt** — telling the AI how to process what you say each time
4. Set a unique **shortcut key combination**
5. Save it. Press-and-hold the shortcut and speak to trigger it.
**Built-in instruction templates (toggle on to use):**
- **Translate to English** — speak in any language, automatically translates to fluent English
- **Terminal Helper** — describe operations in natural language, automatically converts into runnable terminal commands
- **Professional Rewrite** — turns casual speech into polished, professional written language
- **Chaos Mode** — adds emojis and memes, ideal for casual contexts
- **Buzzword Mode** — converts simple ideas into corporate-speak full of industry jargon
**Custom examples:**
> "Format what I say as Notion-style bullet points"
> "Output everything in Slack message style"
> "Organize my thoughts into meeting-minutes format"
---
### Personal Dictionary {#dictionary}
If your work involves proper nouns, brand names, people's names, or industry-specific terms, add them to Speakly's personal Dictionary so the AI can learn them ahead of time and avoid recurring transcription errors.
**Where to access it:** the **Dictionary** page in the left sidebar of the Speakly client.
**Two ways to add words:**
- **Manual add** — open the Dictionary page → switch to the **Manually Added** tab → click **Add** and enter the word you want preserved
- **Auto add** — when you correct a word in a transcription, Speakly will offer to save the corrected spelling to your Dictionary; the next time you say the word, Speakly will output your saved version
**Capacity:** up to 8,000 entries.
**Cross-device sync:** the Dictionary is tied to your Genspark account and syncs across all devices when you're signed in. Dictation settings (shortcuts, custom instructions) sync the same way
**Undo auto-add:** if an automatically added entry isn't right, simply delete it from the Dictionary page.
---
### Dictation settings sync {#dictation-settings-sync}
Your dictation settings now sync across devices — desktop and mobile. Sign in with your Genspark account and your shortcuts, custom instructions, and preferences come with you, so your dictation works the same on every device. That's on top of the existing cross-device sync for the Dictionary.
### Language declaration {#language-declaration}
You can now **declare your language** in Speakly — tell it which languages you primarily speak and it tunes recognition to them, instead of relying on automatic detection alone.
### GenClipboard {#genclipboard}
Need a clipboard manager? **GenClipboard** is Genspark's clipboard tool — searchable copy history, screenshots, and more. See the [GenClipboard article](https://www.genspark.ai/helpcenter/genclipboard).
### Meeting Notes {#meeting-notes}
Beyond real-time dictation, Speakly can record full meetings end-to-end — capturing audio, transcribing, and generating AI summaries. Click **Meeting Notes** in the left sidebar; the top of the page offers three entry points:
**1. New Note (record and save a meeting)**
Click to start recording the current meeting or conversation. Speakly records and transcribes in real time and, when you stop, produces a full transcript and AI summary saved to your Notes list. It recognizes multiple speakers and outputs speaker labels with timestamps.
While recording, you can also open the **Floating Subtitle** window to follow along, pick a **target language** to see both the original and a translation, and switch the **audio source** between microphone and system audio — these share the same real-time engine as Live Translation below.
**2. Live Translation (real-time transcription + translation)**
If you just want to follow along with live subtitles and don't need to save a Note, Live Translation is the lighter option. Click to launch a **Floating Subtitle** window pinned on top of any app showing real-time speech-to-text.
Pick a target language (Chinese, English, Japanese, Korean, Spanish, Portuguese, French, German, Italian, Russian, Arabic, and more — 15+ supported) and Speakly will display both the original transcript and a real-time translation. You can independently toggle the original and the translation — show original only, translation only, or both side-by-side. The audio source also switches between microphone and system audio, making it useful for cross-language meetings, foreign-language lectures, or watching foreign-language video.
**3. Import File (upload audio / video for transcription)**
Already have a recording? Upload it and let Speakly transcribe it. Supported formats: `.m4a` / `.mp3` / `.mp4` / `.aac` / `.wav` / `.ogg` / `.opus` / `.flac`.
Meeting Notes works by capturing system and microphone audio, so it's **not tied to any specific meeting platform**. Use it with Zoom, Teams, Google Meet, or any other video / voice call, as well as in-person conversations, phone calls, podcast recordings, etc.
**Sharing:** Once a transcription is complete, you can generate a share link to send to colleagues.
**Export the raw transcript:** export the full raw transcript in Markdown, Word (.docx), or plain text (.txt), with optional timestamps and speaker labels.
**Note:** Meeting Notes features consume Genspark Credits (billed by duration) and are not covered by Speakly's free quota.
---
## FAQ {#faq data-section-type=faq}
### Use Cases {#usage-scenarios}
#### What is Speakly best for? {#best-use-cases}
The following scenarios deliver the best experience:
- **Email replies** — dictate the email and the AI cleans it up into a professional tone; turning on the Professional Rewrite instruction makes the output even more polished and formal
- **Document writing** — think aloud while you speak, generate drafts quickly, reduce the "staring at a blank page" friction
- **Meeting prep** — speak your thoughts aloud and get an outline or talking points
- **Quick capture** — voice memos without switching apps
- **Multilingual communication** — speak in your native language, get text in the target language
- **Coding assist** — use the Terminal Helper instruction to turn natural language into runnable CLI commands
#### Is it suitable for long, continuous voice input? {#long-dictation}
Yes. Speakly does not impose a hard cap on single-recording duration — as long as you keep speaking or producing audio, the session stays alive. It's well-suited for writing long documents, recording speech drafts, or post-meeting content cleanup. If there's prolonged silence (around 60 seconds without audio), the session may auto-close.
#### Can I customize behavior per app? {#per-app-customization}
Yes. Use the **Custom Instructions** feature to create multiple rule sets and switch between them in different contexts — e.g. use Terminal Helper while coding, Professional Rewrite while writing emails.
---
### Platforms & Compatibility {#platform-compatibility}
#### Which platforms and environments are supported? {#supported-platforms}
Speakly supports both desktop and mobile:
- **Mac / Windows**: full keyboard shortcut experience
- **iPhone / Android**: activated via in-app button; no keyboard shortcuts
Speakly relies on cloud AI for speech recognition and text polishing, so **a network connection is required** — offline use is not supported.
#### Does it work in every app on my computer? {#all-apps}
Yes. Speakly runs as a system-level tool, so it works in virtually any place that accepts text input — including browsers, the Office suite, Slack, various CRM tools, Zoom chat boxes, Teams message bars, and 100+ other applications.
---
### Pricing & Free Trial {#credits-pricing}
#### How is Speakly billed? Is there a free quota? {#pricing-overview}
Speakly's features fall into two categories: **Dictation-class features** are covered by a free quota, while **Agent Mode and Meeting Notes** consume Genspark Credits.
**Dictation-class features (covered by the free quota below):**
- **Default Shortcuts**: Dictation and Hands-free Mode
- **Recommended Shortcuts**: all built-in instruction templates (Translate to English, Terminal Helper, etc.)
- **My Shortcuts**: all your custom instructions
Free quota for Dictation-class features depends on user type:
- **Plus / Pro / Team Plus / Enterprise members**: unlimited use, no word or duration limit
- **First-time users**: automatic **7-day free trial**, unlimited use during this period
- **Invite reward**: for each new user who successfully uses Speakly via your invite, both the inviter and invitee receive **30 days** of free use, stackable on top of existing benefits
- **Free users after the trial**: **4,000 words per week**, resets every Monday at 00:00 UTC
**Important:** the free quotas above (including the 7-day trial and the 30-day invite reward) **only cover Dictation-class features**. Agent Mode (Ask Genspark) and Meeting Notes are not included — using them consumes Credits even during the trial or invite reward period.
**Credit-consuming features (not covered by the free quota):**
- **Agent Mode** (the Genspark Agent tasks summoned by double-pressing the shortcut)
- **Meeting Notes** (covers New Note recording, Live Translation real-time transcription + translation, and Import File transcription)
#### What if I run out of weekly word quota? {#out-of-credits}
When a Free user's weekly quota is depleted, Dictation features will be paused. Options:
- Wait for automatic reset next Monday at 00:00 UTC
- Upgrade to Plus / Pro for instant unlimited use
- Invite a new user to Speakly — once they make their first voice input, both you and the invitee get 30 days of free use
---
### Language Support {#languages}
#### Which languages are supported? {#supported-languages}
Speakly supports 100+ languages, including English, Chinese, Japanese, Korean, Spanish, French, German, Italian, Portuguese, Arabic, Hindi, and many less common languages.
#### Can I mix multiple languages in one sentence? {#multiple-languages}
Yes. Speakly can handle multiple languages mixed in a single sentence, automatically detecting and transcribing each correctly — no manual language switching required.
#### Does it work well for non-English languages? {#non-english}
Yes. Speakly delivers solid recognition quality for major non-English languages, and its AI polish step also cleans up non-English content into well-structured text.
---
### Accuracy & Transcription Quality {#accuracy}
#### Speakly is changing my words or adding text I didn't say — what should I do? {#why-changed}
By default, Speakly applies AI post-processing to clean up transcriptions — removing filler words ("um," "uh"), fixing run-on sentences, and adding punctuation. Occasionally this can over-correct, especially with idioms, proper nouns, or technical jargon.
If you want certain words preserved exactly as you say them:
- Add them to your **Dictionary** — Speakly will always respect your saved words
- Use **Send feedback** to report the specific issue (see below)
#### Why does Speakly mis-transcribe names and technical terms? {#proper-nouns}
Proper names, brand names, and specialized terms are the hardest part of any speech recognition system. Speakly's **Dictionary** feature is designed exactly for this — add names, acronyms, and domain-specific words once, and Speakly will recognize and spell them correctly going forward.
To add a word: left sidebar → **Dictionary** → **Manually Added** tab → **Add**.
#### Does accuracy improve over time? {#accuracy-improves}
Yes. The more words you add to your Dictionary, the better Speakly handles your accent, pace, and vocabulary. Speakly's underlying models also receive regular updates.
#### How do I give feedback on a specific transcription? {#feedback-bad-transcription}
Go to the **Home** tab → scroll to the bottom **Give feedback** section → click the **Last transcript** dropdown to expand the history → select the transcript you want to flag → describe the issue in the input box → click **Send feedback**.
The feedback includes the transcript and the corresponding audio file to help improve recognition quality. Speakly does **not** use this data to train general models.
---
### Technical Issues {#technical-issues}
#### My shortcut isn't working — what's wrong? {#shortcut-not-working}
Run through these checks in order:
1. Confirm the Speakly app is running in the background
2. Check that the system **Accessibility** permission is granted to Speakly (Mac: System Settings → Privacy & Security → Accessibility)
3. Make sure no other app is using the same shortcut
4. Restart Speakly and try again
#### The app is unresponsive — what should I do? {#app-not-responding}
1. Fully quit and relaunch Speakly
2. Check that your network connection is working
3. If the issue persists, uninstall and reinstall the latest version
#### Speakly didn't insert text at my cursor position — what's wrong? {#wrong-insertion}
If text isn't going to your cursor, the focus probably shifted away during recording, or system Accessibility permission isn't properly granted.
Steps to check:
1. Verify Accessibility permission: Mac → System Settings → Privacy & Security → Accessibility → ensure Speakly is checked
2. Keep the target app focused throughout the recording — don't switch to other windows mid-dictation
3. If insertion still fails, Speakly will automatically show a popup with the transcribed text so you can copy and paste manually
#### Where can I see my transcription history? {#transcription-history}
Open the **Home** tab to see your full transcription history, grouped by Today / Yesterday / etc. You can search by keyword to find specific transcripts.
#### How do I disable custom instructions for a specific task? {#disable-custom-keys}
In Speakly's settings or instruction management page, you can temporarily disable a specific custom instruction or switch to the basic transcription mode (no instruction applied).
#### Is what I say being saved? Is the data secure? {#privacy}
Speakly transmits and processes voice data over encrypted channels and does not permanently store your raw audio. For full details, see the [speakly.ai Privacy Policy](https://speakly.ai/privacy).
#### How do I report issues or send feedback? {#how-to-feedback}
You can submit feedback directly inside the Speakly client:
1. Switch to the **Home** tab
2. Scroll to the bottom of the page to find the **Give feedback** section
3. Click the **Last transcript** dropdown to expand the history, then pick the transcript you want to flag (defaults to the most recent one)
4. Describe the issue or suggestion in the **How can this transcript be improved?** input box
5. Click **Send feedback** to submit
#### I speak more than one language — can I tell Speakly? {#multilingual-declaration}
Yes. Declare your languages in Speakly and recognition is tuned to them. You can still mix languages in one sentence; auto-detection keeps working where you don't declare.
#### Do my Speakly settings sync between my computer and phone? {#settings-sync}
Yes. Dictation settings sync across your devices when you're signed in with your Genspark account, alongside the Dictionary.
# GenTerminal {#genterminal data-doc-group=build-suite}
Source: https://www.genspark.ai/helpcenter/genterminal
## What is GenTerminal? {#what-is-genterminal data-section-type=overview}
**GenTerminal** is Genspark's cross-platform SSH terminal client for **iPhone, Android, Mac, and Windows**. Connect to your servers, manage your hosts, and run commands from any device — with your hosts and settings kept in sync through your Genspark account.
Core capabilities:
- **SSH terminal** — connect to any server over SSH and work in a full-featured terminal, on desktop or mobile
- **Host management** — save your servers, organize them into groups, and reconnect to recent hosts in one tap
- **Cross-device sync** — sign in with your Genspark account and your saved hosts and settings follow you to every device
- **Encrypted key vault** — protect your saved passwords and private keys with a vault password
- **SFTP file transfer** — browse, upload, and download files on your server
- **Port forwarding** (Mac & Windows) — set up local and remote tunnels
- **Snippets** — save commands you reuse and run them with one tap
- **Quick window toggle & system tray** (Mac & Windows) — summon or hide the window with a global hotkey and keep it out of the way in the tray
---
## Quick Start {#quick-start data-section-type=overview}
1. **Download and install GenTerminal**
- **iPhone** — get it from the App Store
- **Android** — get it from Google Play
- **Mac / Windows** — download the desktop app from [genterminal.ai](https://genterminal.ai/)
Then open the app.
2. **Sign in with your Genspark account**
Sign in with your existing Genspark account. GenTerminal is integrated with Genspark, so no separate registration is required — and signing in is what enables your hosts to sync across devices.
3. **Add a host**
Tap or click **Add Host** (the **+** button) and fill in your server details:
- **Host / address** — the server's hostname or IP address
- **Port** — the SSH port (usually `22`)
- **Username** — your login user on the server
- **Authentication** — either a **password** or a **private key**
Save the host.
4. **Connect**
Open the host you just saved to start an SSH session. On Mac and Windows you type commands with your keyboard just like any terminal; on iPhone and Android, use the function-key bar above the keyboard for keys like Tab, Ctrl, Esc, and the arrows.
---
## Feature & How-To Guides {#feature-guide data-section-type=feature}
### Host Management & Groups {#host-management}
The host list is your home base. From there you can:
- **Add, edit, and delete hosts** — keep all your servers in one place
- **Organize into groups** — sort related servers (for example by project or environment) into groups so the list stays tidy
- **Reconnect to recent hosts** — your most recently used connections are surfaced for one-tap reconnect
---
### Cross-Device Cloud Sync {#cloud-sync}
When you're signed in with your Genspark account, GenTerminal keeps your hosts, groups, snippets, and settings in sync automatically across all your devices. Add a server on your phone and it's there on your Mac too — no manual export or import.
Sync is tied to your Genspark account, so signing in on a new device restores everything you've saved.
---
### Key Vault {#key-vault}
The Vault protects the sensitive parts of your hosts — saved passwords and private keys — by encrypting them with a **vault password** that only you know.
- **Set a vault password** — the first time you save credentials, you'll be asked to create a vault password. This password unlocks your saved keys on any device.
- **Unlock on a new device** — entering the same vault password on another device decrypts your synced credentials there. Your vault password is never stored on our servers, so keep it somewhere safe — it can't be recovered for you.
- **Encrypt keys before sync** — a setting that encrypts your private keys with your vault password *before* they leave your device, so only ciphertext is ever synced. Turn it on if you want your keys protected end-to-end in sync.
---
### SFTP File Transfer {#sftp}
Beyond running commands, GenTerminal can browse the files on your server over SFTP. Open the file browser for a connected host to navigate folders, **upload** files from your device, and **download** files from the server.
---
### Port Forwarding {#port-forwarding}
On **Mac and Windows**, GenTerminal can set up SSH tunnels so you can reach a service through your server:
- **Local forwarding** — make a port on a remote network reachable from your own machine (for example, to open a database or internal web app that only your server can see)
- **Remote forwarding** — expose a port from your machine on the remote side
Set the tunnel up alongside the host connection and it stays active while you're connected.
---
### Snippets {#snippets}
Snippets are commands you save once and reuse. If you find yourself typing the same command often, save it as a snippet and insert it into the terminal with a tap or click instead of retyping it.
---
### Quick Window Toggle & System Tray {#quick-toggle}
On **Mac and Windows**, GenTerminal stays handy without cluttering your screen:
- **Global hotkey** — press a keyboard shortcut from anywhere to instantly show or hide the GenTerminal window. You can change the shortcut in **Settings**.
- **System tray** — closing the window keeps GenTerminal running in the tray (menu bar on Mac), so your sessions and quick toggle stay available. Quit fully from the tray menu when you're done.
---
### Function-Key Bar {#function-keys}
On **iPhone and Android**, phone keyboards don't have the keys a terminal needs, so GenTerminal adds a function-key bar above the keyboard with **Tab**, **Ctrl**, **Esc**, the **arrow keys**, and other special keys — so you can run interactive tools and shortcuts just like on a desktop terminal.
---
## FAQ {#faq data-section-type=faq}
### Which platforms does GenTerminal support? {#supported-platforms}
GenTerminal runs on **iPhone, Android, Mac, and Windows**. Get the mobile apps from the App Store and Google Play, and the Mac and Windows desktop apps from [genterminal.ai](https://genterminal.ai/). Your hosts and settings sync across all of them when you're signed in with the same Genspark account. Port forwarding and the global hotkey / system tray are available on the desktop apps; the function-key bar is on mobile.
### On iPhone, my SSH session disconnects shortly after I switch away from the app — why? {#ios-background-disconnect}
This is expected behavior on iPhone, not a bug. iOS limits what apps can do in the background, so after GenTerminal has been in the background for about 30 seconds, the operating system suspends it and the SSH connection drops. When you return to the app, simply reconnect to the host. To avoid an interruption during a long-running command, keep GenTerminal in the foreground. (The Mac and Windows apps are not affected by this limit.)
### How do I send diagnostic information when I report a problem? {#send-diagnostics}
If you hit an issue and our support team asks for logs, you can upload diagnostics from inside the app:
1. Open **Settings**
2. Tap or click **Send diagnostics**
This uploads recent app logs so we can investigate. Mention the approximate time the problem happened when you contact support — it helps us find the right logs.
### How do I get support? {#contact-support}
Email us at **support@genspark.ai**. Include what happened, which device and OS version you're on, and — if you've already uploaded diagnostics — the approximate time of the issue.
### How do I delete my account? {#delete-account}
You can delete your account from inside the app. Open **Settings**, find the account section, and choose **Delete account**. Deleting your account removes your synced data; this action cannot be undone.
### Do my saved hosts sync across devices? {#sync-devices}
Yes. As long as you're signed in with the same Genspark account, your hosts, groups, snippets, and settings sync automatically across iPhone, Android, Mac, and Windows. To decrypt synced passwords and keys on a new device, enter your vault password there.
# AgentBase {#agentbase data-doc-group=build-suite}
Source: https://www.genspark.ai/helpcenter/agentbase
## What is AgentBase? {#what-is-agentbase data-section-type=overview}
**AgentBase** is Genspark's AI-powered business system builder — it turns one sentence into custom **Dashboards, CRM, and business systems** in minutes. Describe what you want to manage, and AI builds the database, fills in starter data, and generates a ready-to-use view. No spreadsheets, no schema design, no SQL required.
Standing up a business system traditionally means figuring out which tables you need, designing fields, entering data row by row, and manually configuring dashboards. AgentBase collapses all of that into **one sentence → a working system**, with no technical knowledge required.
### Primary Use Cases {#use-cases}
- **CRM** — customers, contacts, deals, follow-up notes
- **Inventory & orders** — products, stock levels, incoming/outgoing
- **Content calendar** — topics, status, platforms, schedule
- **Recruiting pipeline** — candidates, roles, interview stages, evaluations
- **Personal trackers** — reading notes, budgets, travel plans, goals
### Who It's For {#who-its-for}
Anyone who needs to manage "a structured set of items" over the long term — sales teams, operations, HR/recruiting, marketers, small-business owners, and individuals — **without** wanting to learn a database tool.
### Key Benefits {#key-benefits}
- **One sentence → working system** — tables, data, and views all AI-built
- **Persistent and reusable** — your workspace stays; keep adding data, switching views, and sharing over time
- **No technical skill needed** — fully conversational; you never touch schemas or query syntax
- **Multi-source consolidation** — pull data from files, email, apps, and databases into one place
- **From insight to action** — draft follow-ups, flag risks, trigger automated reports
## Quick Start {#getting-started data-section-type=overview}
No installation required. All you need is a signed-in Genspark account.
### Quick Start — Your First System in Under 10 Minutes {#quick-start}
1. Navigate to the Genspark homepage and click the **AgentBase** icon.
2. **Choose your starting point:**
| Option | When to Use |
|---|---|
| **Just describe it** | Start from a plain-language prompt — AI builds from scratch |
| **From files** | Upload CSV, Excel, PDF, or Word |
| **From your inbox** | Pull in Gmail or Outlook data |
| **From another app** | Connect HubSpot, Salesforce, or a database |
| **Template Center** | Browse by category tab → click a card → preview → **Use this template** |
3. **Describe what you want.** Type one sentence in the input box, for example:
```prompt
Help me manage customers and sales deals
```
```prompt
Set up a recruiting pipeline to track candidates and interview stages
```
```prompt
Create a vendor management system with contract renewal dates
```
4. **Verify your workspace.** You'll know it worked when:
- The left sidebar shows **Dashboards / Tables / Workflows** groups
- The right pane shows your generated tables and a first view with starter data
- Your system appears under **My Systems** on the AgentBase homepage
From here, keep chatting to refine the structure, or move straight to importing your real data.
## Core Concepts {#core-concepts data-section-type=overview}
### The Three-Layer Model {#three-layer-model}
Everything in AgentBase maps to three layers. Build them in this order:
**Tables (data layer) → Views & Dashboards (presentation layer) → Workflows (automation layer)**
Get the data right first, shape the views second, automate third.
### Key Terms {#key-terms}
| Term | Definition |
|---|---|
| **Workspace / System** | A persistent container for one business domain (e.g. "Sales CRM"). Holds all tables, views, dashboards, and workflows. |
| **Table** | A relational data table built by AI. Multiple tables can link to each other (e.g. "Customers" → "Deals"). |
| **View** | A way of presenting your table data — Table / Kanban / Gallery / Calendar / Form. Switching or deleting a view never changes the underlying data. |
| **Dashboard** | A summary page pulling KPI cards and charts from multiple tables, with clickable filters. |
| **Workflow** | A schedule-triggered automation that syncs data or sends reports in the background. |
| **Starter data** | Sample records AI seeds so you can see the structure immediately. Replaced when you enter real data. |
| **Aggregated view** | For very large datasets — AI imports summary results only (not every row) to keep performance fast. |
| **Role** | Access level on a shared workspace: Viewer / Editor / Owner. |
### Foundational Rules {#foundational-rules}
- **Everything is conversational** — describe intent; AI handles the technical work
- **One system per conversation** — to build or switch to another system, return to the homepage or start a new conversation
- **Views are non-destructive** — switching or deleting a view never touches the underlying data
- **Persistence is the point** — your system stays around for continuous use, unlike a one-off answer or spreadsheet
## Feature & How-To Guides {#feature-guide data-section-type=feature}
### Importing Your Data {#importing-data}
All imports happen through chat — tell AI where your data is. No manual upload screens or connection wizards necessary.
| Source | How to Connect |
|---|---|
| **Files** | Upload Excel / CSV / PDF / Word, or a whole folder (Browse Local Files or Choose from AI Drive) |
| **Email** | Gmail or Outlook — one-time authorization on first use |
| **Microsoft 365** | SharePoint, OneDrive, Teams, Outlook calendar & contacts |
| **HubSpot** | Paste a **Private App Access Token (PAT)** generated in HubSpot — this is not a one-click admin authorization |
| **Salesforce & others** | Follow the UI prompts for authorization |
| **Databases** | PostgreSQL, MySQL, Snowflake — provide connection info when prompted |
**Large datasets:** AI automatically builds an *aggregated view* for very large files — it imports summary results rather than every row to keep performance fast. Use filter prompts to work with subsets.
### Views & Dashboards {#views-dashboards}
Once your data is in, switch between views based on what you're trying to accomplish. Switch via the toolbar buttons (instant, always free) or ask AI in chat.
| View | Best For |
|---|---|
| **Table** | Full-field browse and bulk editing — your source of truth |
| **Kanban** | Process / stage tracking (deals, hiring pipeline, task triage) |
| **Gallery** | Image-heavy or card-based content (products, assets, profiles) |
| **Calendar** | Date-based data (renewals, content schedule, interview dates) |
| **Form** | Quickly adding new records via a shareable intake form |
| **Dashboard** | KPIs, totals, trend charts, and risk highlights across multiple tables |
**How to build a dashboard:** Describe what you want to track in chat, for example:
```prompt
Create a dashboard showing total contract value by vendor, renewals in the next 90 days, and any accounts flagged as at-risk
```
AI generates KPI cards and charts. Click any chart element to filter the underlying table. Viewing a finished dashboard is always free.
### Querying Your Data {#querying-data}
No SQL or formula knowledge needed — ask in plain language and AI returns answers with cited cell references so you can verify the source.
```prompt
Which records need action this week and why?
```
```prompt
List products with stock below 10
```
```prompt
Show accounts with no activity in the last 30 days
```
### Workflows & Automation {#workflows-automation}
Workflows let you automate recurring actions on a schedule — syncing data, sending reports, or flagging items that need attention.
To create a workflow:
1. Once your views look correct, describe the automation in chat, for example: *"Every Monday morning, compile last week's new deals into a summary and email it to me."*
2. AI builds the workflow with a schedule (Daily / Weekly / Monthly / Custom + timezone) and shows a **Next Run At** timestamp.
3. Run a **Test** before enabling — AI simulates the trigger without firing real actions.
4. Toggle the workflow **On** when you're satisfied with the output.
**Best practice:** Always set new communication workflows to **draft-only** first. Review the wording and trigger logic before enabling auto-send. Never auto-send to customers until you fully trust the output.
## Sharing & Collaboration {#sharing-collaboration data-section-type=feature}
Share your workspace, a single dashboard, or a single table with other people, and control what each person can do.
### Choose what to share {#choose-what-to-share}
When you open the **Share** dialog, you can set the scope of the share:
- **Whole workspace** — the person gets access to everything in the workspace (the original behavior, and still the default).
- **A single dashboard** — the person sees only that dashboard.
- **A single table** — the person sees only that table.
Sharing is controlled by who you invite and what role you give them — not by the URL. Everyone with an active invite sees the same scope you set, regardless of how the link is shared.
### Invite people and assign a role {#invite-people-and-assign-a-role}
In the Share dialog, enter the person's email address, select a role, and click **Invite**. They'll receive an invitation and get access at the scope you set. Choose their level of access:
| Role | What they can do |
| :---- | :---- |
| **Viewer** | View data only |
| **Editor** | View and edit data |
| **Owner** | Full control, including managing sharing |
## Troubleshooting {#troubleshooting data-section-type=troubleshooting}
### Common Problems & Solutions {#common-problems}
| Problem | Solution |
|---|---|
| Report email never arrived | Connect Gmail/Outlook — the send step fails without a linked mailbox. Connect it and the workflow delivers on the next run. |
| Can't find a "Blank Workspace" card | There isn't one — use **Just describe it** or type directly into the composer. |
| No search box in Template Center | Browse via the left category tabs — there is no search box. |
| The table structure AI built isn't right | Tell AI in chat: *"Add a column for X"* / *"Delete this table"* / *"Change what this field means."* |
| Large dataset slowed or wasn't fully imported | AI builds an aggregated view automatically for large files — use filter prompts to work with subsets. |
| Workflow ran but nothing happened / errored | Check the run history for the error and confirm the mailbox is connected for email steps. Fix the existing workflow in place. Ask the agent to change it, or open the builder; edits apply to the same workflow object so run history and schedules survive. |
| Shared link shows "no access" | Check (1) active invite exists (2) invite covers the scope they're opening (3) signed in with the account the invite was sent to.. |
| Dashboard totals look wrong | Validate the source table first — check field mapping, date formats, and blank/null values before trusting dashboard output. |
| Fields mapped incorrectly after import | Use a follow-up prompt: *"Rename 'Col A' to 'Account Name' and map the date column to 'Renewal Date'."* |
## FAQ {#faq data-section-type=faq}
### Do I need technical skills to use AgentBase? {#need-technical-skills}
* No. The entire experience is conversational — you never touch schemas, queries, or database configuration of any kind.
### How is this different from AI Sheets? {#vs-ai-sheets}
* AI Sheets is a single table with calculations. AgentBase is a multi-table relational system with views, dashboards, workflow automation, and team sharing — built for managing a business category long-term, not just crunching numbers in one tab.
### Will my system stay around? {#persistence}
* Yes. Every workspace is persistent and listed under **My Systems** on the homepage. Come back anytime to add data, switch views, or share.
### Will scheduled workflows run if my computer is off? {#workflows-offline}
* Yes — workflows run in Genspark's cloud, completely independent of your device.
### Can I use this instead of Salesforce or HubSpot? {#vs-salesforce-hubspot}
* Yes for core CRM use cases — pipeline tracking, account management, deal dashboards. It can also pull data *from* Salesforce or HubSpot to create a simplified view alongside other sources.
### Is this only for sales teams? {#not-only-sales}
* No. It works for any team managing structured data — HR pipelines, vendor contracts, project tracking, content calendars, onboarding checklists, and more.
### Can multiple people use the same workspace at once? {#multiple-people}
* Yes. Invite teammates via the Share dialog and assign roles. Editors can update data and views; Viewers can browse and read reports.
### Can I upload customer phone numbers and email addresses to AgentBase? Is it safe? {#customer-data-upload}
* You may use AgentBase to store and manage customer contact data, such as phone numbers and email addresses, provided that your business has a lawful basis and all necessary rights or permissions to collect, use, and process that data.
* Genspark applies security and access controls to stored data, including encryption and protections against using stored data to train generalized AI models.
### If my business collects customer personal data, do I need separate consent for the cross-border transfer to Genspark? {#cross-border-data-transfer-consent}
* This depends on the laws that apply to your business and the individuals whose data you collect. You are responsible for complying with applicable requirements for collecting, using, disclosing, and transferring personal data, including providing any required notice, obtaining consent where required, and using any required cross-border transfer mechanism.
* A privacy-policy disclosure may be sufficient in some circumstances, but it does not automatically replace consent or other legally required safeguards. Businesses should consult qualified legal counsel if they are unsure what requirements apply.
* Any content entered into AgentBase—including prompts, uploaded files, and customer data—is the business’s responsibility under the applicable Genspark terms and policies. See the [Genspark Privacy Policy](https://www.genspark.ai/privacy) and [Terms of Service](https://www.genspark.ai/terms).
# Genspark Design {#genspark-design data-doc-group=build-suite}
Source: https://www.genspark.ai/helpcenter/genspark-design
## What is Genspark Design? {#what-is-genspark-design data-section-type=overview}
Genspark Design is an AI-powered creative suite that takes you from a rough idea to a professional, production-ready design — entirely through conversation. No design software skills required.
Describe what you need in plain language, and Genspark Design handles the concept, layout, color, imagery, and interactivity. Whether you're a product manager sketching a prototype, a marketer creating event posters, or a founder building a landing page, the AI does the heavy lifting while you stay in control.
- **What you can build:** UI prototypes · websites · posters · animations · presentations · marketing assets · documents
- **Who it's for:** Product managers, marketers, founders, engineers, and creators who need professional design output without Figma or Canva expertise.
- **Try it now:** Open Genspark Design.
## Quick Start {#getting-started data-section-type=overview}
1. Describe your design in the input box — be as specific or open-ended as you like.
2. Optionally attach context via the **+** button: a Figma file, screenshot, GitHub repo, or a Design System.
3. Send — your design is generated at the highest quality by default. The AI may ask clarifying questions, but it may also start straight away — include the size, colors, type, and style you want up front.
**Tip:** Not sure where to start? Click any Popular Tasks example below the input box — it fills a ready-made prompt you can edit before sending.
## Core Concepts {#core-concepts data-section-type=overview}
### Key Terms {#key-terms}
| Term | What it means |
| --- | --- |
| Canvas | Right-side panel where your design lives and is edited |
| Chat panel | Left-side panel where you talk to the AI |
| Design System | Saved brand style (colors, fonts, components) applied automatically to every design |
| Tweaks | Live parameter controls for real-time fine-tuning — no re-prompting needed |
| Build it | One-click export of your design as a runnable website or app |
| Verifier | An optional quality check the agent can run in the background; it only messages you when it finds something to fix |
| Remix | Copy a shared design to create your own editable version |
| History panel | A list of earlier versions of a project |
### Design Types {#design-types}
Just describe what you want; the AI picks the right type automatically.
| Type | Best for |
| --- | --- |
| Prototype | Interactive apps, UX flows, SaaS dashboards |
| Slide deck | Pitch decks, reports, training materials |
| Poster | Events, social media, print assets |
| Video / Frame Movie | Product intros, logo animation, multi-shot ads |
| Wireframe | Low-fidelity layout exploration |
| Flipbook / Paged Book | Lookbooks, whitepapers, technical manuals |
### What Design Can Connect To {#what-design-can-connect-to}
Genspark Design has its own toolset: it can connect to GitHub, browse a repository, import files, work with Figma, capture elements from a web page, and search images. Two things to keep in mind:
- Attaching a repository as context gives the agent a snapshot of the files you attach — small text and code files, filtered and capped — for that message. It isn't a live connection to your whole repo, and the agent doesn't run your code.
- Using a tool — for example, importing files from a repo — brings those files into the project.
For coding work in a full developer environment, use [Genspark Code](https://www.genspark.ai/helpcenter/ai-developer).
## What Design Costs {#what-design-costs data-section-type=feature}
Genspark Design is billed in credits.
- **Credits only — not Free Quota.** Genspark Design isn't covered by Free Quota on any plan. For what Free Quota covers, see [Membership Plans](https://www.genspark.ai/helpcenter/membership-plans).
- **Billed as the agent works — not by page or unit.** Credits are used as the agent builds the design, and each image, video or audio generation is charged on top — so asking for lots of images is a common way a project gets expensive.
- **Retries and redraws cost again.** Re-running a change or redrawing is a new, billed generation, so a back-and-forth fix loop adds up.
- **Where to see it:** open Credit Usage → Design in your account. For how credits are counted and charged, see the [Credits Guide](https://www.genspark.ai/helpcenter/credits-guide).
## Feature & How-To Guides {#feature-guide data-section-type=feature}
### Creating a Design {#creating-a-design}
- **From scratch** — type a description, attach optional context, then send.
- **From a template** — browse the For You gallery → click a design → Remix.
- **Multiple directions** — open multiple chats within one project (each has independent memory).
**Adding context to your prompt:** Click the **+** button in the input box to attach supporting materials — local files, images, or docs; files from your AI Drive; a Figma file; a GitHub repository; a reference to one of your past designs; a web element grabbed from any webpage; or a saved Skill or Design System. The more context you give upfront, the closer the first generation will be to what you need.
### Editing on the Canvas {#editing-canvas}
| Tool | How to use | Best for |
| --- | --- | --- |
| Tweaks | Adjust sliders and color pickers in real time | Global parameters: colors, spacing, text |
| Edit | Click any element, then modify it in the panel | Changing one specific element. Edit targets the element you select; simple text and style changes usually apply directly. If a change can't be applied directly, it's handled as a new chat generation, which can re-bill and may affect other elements. |
| Comment | Pin feedback on the Canvas, then send it all at once | Descriptive feedback ("change this to blue") |
| Draw | Sketch with the red brush or add sticky notes, then send | Visual instructions: arrows, circled areas |
### Edit Safety: Protect Your Work {#edit-safety-protect-your-work}
A common frustration is a small change that redraws work you wanted to keep. A few habits prevent it:
- Change one thing at a time. Small, separate edits are easier to keep than one big rework.
- Use Edit for a single element, and Comment to pin several notes and send them together.
- Simple text and style changes usually apply directly — often with no redraw and no extra credit cost.
- If a change can't be applied directly — or it's a layout, structure or new-element change — it's handled as a chat generation. That re-bills and may affect other elements, so check the result before continuing.
- Duplicate the project before a big change, so the version you like stays intact.
- The History panel is read-only. It shows earlier versions, but it can't roll back to one.
### Design System {#design-system}
1. Click **Design System** below the input box → **Create New**.
2. Enter a brand description; import assets (Figma .fig / GitHub / codebase / uploaded files).
3. AI extracts colors, fonts, and components — review and confirm.
4. Reuse it across every project. You can also extract a Design System from any existing design via the Share menu → Create design system from this project.
### Exporting {#exporting}
| Output | How to trigger | What it's for |
| --- | --- | --- |
| Present | Click Present in the top bar | Fullscreen playback; arrow keys to flip pages; Speaker Notes for slide decks |
| Build it | Click Build it → Website / App | A runnable website or app. It stays inside Genspark Design — a developer handoff is a separate route |
| ZIP (project source) | Click Download — its own button next to Build it in the top bar; disabled while a generation is running | The project's source files — not just a developer handoff document |
| PNG | Export menu → Export image (poster designs) | A raster image |
| PSD | Export menu → Export PSD (poster designs) | A layered PSD — one Photoshop layer per design layer. Some effects may not rasterize; the export still downloads, minus those layers |
| MP4 | Export settings dialog | Video |
| PPTX | Say "Export as PPTX" in chat | Editable slides. Styling is approximate — effects PowerPoint can't represent are dropped |
| PDF | Say "Export as PDF" in chat | A print-ready file, produced through your browser's print dialog |
The Export menu with PNG and PSD appears on poster designs. For PDF and PPTX behavior, see [Export Fidelity & Troubleshooting](#export-fidelity-and-troubleshooting).
### Export Fidelity & Troubleshooting {#export-fidelity-and-troubleshooting}
- PDF is rendered by your browser, not by Genspark. The auto-print pop-up is often blocked — allow pop-ups and try again, or use the download link the agent posts. Background colors can be dropped depending on your print settings.
- PPTX styling is approximate — that's expected. Complex effects (gradients, filters, blend modes, custom shapes) aren't representable as PowerPoint objects and may be dropped.
- No vector output. There is no SVG, Adobe Illustrator (.ai), or EPS export; the PSD is layered raster.
- If an export looks blank or fails, try once more. Rendering can miss an asset on the first pass.
### Sharing & Collaboration {#sharing}
- **Share link** — Share icon in the top bar → Anyone with the link (public) or Restricted (email invite).
- **Roles** — Viewer / Editor / Owner.
- **Duplicate** — a full copy of the project including files and conversations.
- **Add to Hub** — the ⋯ menu in the top bar → adds the project to your Hub, a shared team workspace with persistent context.
- **What a public link exposes** — A public link lets anyone with the URL view the design and its source files, version history, and downloads — no sign-in required.
- **Revoke a link** — To cut off access, set the project back to private (Restricted); anyone using the link then loses access.
### Managing Your Projects {#managing-your-projects}
- Find your designs in the My Designs gallery on the Design home page. Click a design to reopen it.
- Duplicate a project from its menu to keep a copy you can experiment on.
- Remix a shared design to create your own editable version.
- Rename or delete a project from its menu in the My Designs gallery. Deleting a project can't be undone.
## Quick Reference {#reference data-section-type=feature}
### Canvas Toolbar {#canvas-toolbar}
| Location | Controls |
| --- | --- |
| Top-right | Tweaks · Edit · Comment · Draw |
| Top bar | Present · Build it · Download · Share · ⋯ menu (Bookmark, Add to Hub) |
| Canvas tabs | Design System |
## Troubleshooting {#troubleshooting data-section-type=troubleshooting}
### The Tweaks panel is empty — what do I do? {#tweaks-empty}
Say "Make this tweakable" in the chat — the AI will generate controls automatically.
### Build it is grayed out — why? {#build-it-grayed-out}
Build it activates once generation is complete. It's disabled while a generation is running, and it isn't available for every design type. Wait for the task list to finish; if it stays grayed out, the design type may not support Build it.
### It generated something I didn't ask for. {#unexpected-output}
Add more detail to your prompt; specify the design type, target audience, dimensions, or visual style. Attach reference screenshots. Use the Comment or Draw tools to point to exactly what needs changing.
### A run fails, freezes, or comes back blank. {#a-run-fails-freezes-or-comes-back-blank}
Try these in order:
1. Reload the project and check whether the run actually finished — the canvas can lag behind the chat.
2. Ask again. A new run is a new charge.
3. Duplicate the project to keep a clean copy before you retry.
Still stuck? Report it with the project link: open the project, click Share, and invite support@genspark.ai as a Viewer.
### An export is missing, blank, or won't open. {#an-export-is-missing-blank-or-wont-open}
- **Missing from the menu:** some formats depend on the design type — the Export menu with PNG and PSD appears on poster designs.
- **PDF or PPTX problems:** see [Export Fidelity & Troubleshooting](#export-fidelity-and-troubleshooting).
- **A format is locked:** export availability can vary by plan.
For more known issues and fixes, see Common Issues.
## FAQ {#faq data-section-type=faq}
### I want to undo a change. {#undo-a-change}
Genspark Design has no one-click undo. The History panel shows earlier versions, but it's read-only — you can look, not roll back. Asking the AI to revert starts a new, billed generation, and it may not reproduce the exact previous state. Make changes one at a time, and Duplicate the project when you reach a version you want to keep.
### Can the code from Build it go straight to production? {#build-it-production}
It runs as-is, but have a developer review it before deploying.
### How is this different from AI Slides? {#vs-ai-slides}
Genspark Design is a full creative canvas (websites, apps, videos, posters, code). AI Slides is purpose-built for presentations with PPTX / Google Slides export. For pitch decks, either works. For prototypes, visuals, or video — use Genspark Design.
### Can I use my designs commercially? {#can-i-use-my-designs-commercially}
Genspark Design is intended for personal and commercial work. Ownership, commercial use, and what you can create are covered by our policies.
### Why did other parts of my design change when I only asked for one edit? {#why-did-other-parts-of-my-design-change-when-i-only-asked-for-one-edit}
Small edits (text, color, style) are written straight into your design. Big edits (layout, structure, new elements) are sent to the AI as a prompt, so the AI rebuilds that part — and may change more than you asked. To stay safe: one change at a time, use Edit for a single element, and Duplicate before a big change.
### Why was I charged for edits I didn't keep, or for a generation that failed? {#why-was-i-charged-for-edits-i-didnt-keep-or-for-a-generation-that-failed}
Design bills for the work the AI does, not for what you keep — so a discarded redraw or a retry still uses credits. Generations blocked before they start are refunded automatically.
### Can I edit a design without the AI regenerating it? {#can-i-edit-a-design-without-the-ai-regenerating-it}
Partly. Simple text and style/Tweaks edits are applied directly — no redraw, no credit cost. There is no full manual-edit mode and no element locking; anything else goes through the AI as a generation.
### Why doesn't my PDF or print export match the design? {#why-doesnt-my-pdf-or-print-export-match-the-design}
PDF exports come from your browser's print dialog, so they follow your print settings — and browsers block the pop-up or leave out background colors by default. PPTX is approximate, since PowerPoint can't represent every effect. Neither export is an exact copy of the design.
### What does a share link expose? {#what-does-a-share-link-expose}
A public link grants read to anyone with the URL.
### What's included in my plan, and do the promos cost credits? {#whats-included-in-my-plan-and-do-the-promos-cost-credits}
Design is on every plan and is credit-only — not covered by Free Quota.
### Can Design use my GitHub repo or the tools Super Agent can? {#can-design-use-my-github-repo-or-the-tools-super-agent-can}
Yes — Design works with GitHub, Figma, web capture, and image search. Connecting GitHub gives it persistent repo tools; attaching a repo with + attaches a snapshot of small files to that message only. Either way, Design reads your code, it doesn't run it.
### How do I control how many images it generates? {#how-do-i-control-how-many-images-it-generates}
There's no cap and no approval step — Design decides how many images to make, and may make more than you asked for. State an exact number, start with a small sample set, then request the rest once you've picked a direction. These are prompt instructions, not enforced limits.
### How do I get better-looking output? {#how-do-i-get-better-looking-output}
Give it a Design System, attach reference images or a Figma file, and describe the design type, size, style, and audience before you send. Use Comment or Draw to point at exactly what to change, and refine one element at a time.
# Skills {#skills data-doc-group=custom-agents}
Source: https://www.genspark.ai/helpcenter/skills
## What Are Skills? {#what-are-skills data-section-type=overview}
**Skills** are reusable AI tools built for specific jobs. Instead of rewriting the same instructions every time, you save a proven workflow once and apply it with one click — consistently, across any project.
A Skill is more than a template. It packages expert thinking, structure, and design so every output follows the same quality bar and reasoning, every time. A sales proposal Skill doesn't just look like a sales deck — it thinks like a sales director, structuring the argument, anticipating objections, and framing the value proposition.
### Key Benefits {#key-benefits}
| Benefit | What It Means for You |
|---|---|
| **Build once, reuse forever** | Define a workflow once and apply it in one click from anywhere |
| **Consistent output quality** | Every result follows the same structure, tone, and expert reasoning |
| **Community library** | 100+ public Skills available to add with one click — no building required |
| **Works across products** | Skills enhance AI Slides, AI Docs, Super Agent, and more |
### Who Skills Are For {#who-its-for}
Skills are for anyone who does recurring, structured work and wants more consistent results with less effort — including:
- Business and operations professionals building the same reports, proposals, or briefs regularly
- Sales and marketing teams producing pitch decks, proposals, and campaign briefs to a standard
- Researchers and analysts running repeatable data analysis or literature review workflows
- Product and engineering teams generating PRDs, release notes, and technical specs
- Team leads who want consistent output quality across a team without requiring everyone to write expert-level prompts
### Skills vs. Templates {#skills-vs-templates}
| Capability | Template | Skill |
|---|---|---|
| Visual style | Yes | Yes |
| Content structure | No | Yes |
| Expert reasoning | No | Yes |
| Reusable prompts | No | Yes |
| Works across products | No | Yes |
A Skill works like a packaged expert, while a template only reuses a layout.
## Quick Start {#getting-started data-section-type=overview}
### Prerequisites {#prerequisites}
- A Genspark account (free or paid)
- No technical background required
### Quick Start {#quick-start}
1. **Browse Skills.** From the Genspark homepage, click **Skills** in the left sidebar.
2. **Apply or create a Skill.** Click any Skill card to preview it, then click **Add & Use** to apply it. Or create your own by clicking **+ New Skill**.
3. **Get expert-quality results.** AI follows the Skill's approach — structure, reasoning, and style — to produce consistent output. Review, refine, and export when ready.
Sample prompts to try:
```prompt
Create a 10-slide investor pitch deck for a Series A SaaS startup
```
```prompt
Write a competitive analysis report comparing the top 5 CRM platforms
```
```prompt
Create a call prep brief for an upcoming sales meeting, including company background, recent news, likely priorities, and suggested talking points
```
### When to Use a Skill {#when-to-use}
| Option | Best for |
|---|---|
| **Standard prompt** | One-off tasks you won't repeat |
| **Existing Skill** | Repeated work where a proven structure already exists |
| **New Skill** | Recurring work that isn't yet standardized, or work to share with a team |
| **AI Slides + Skill** | Presentations needing repeatable slide structure and visual consistency |
### The Three Sources of Skills {#skill-sources}
| Source | Description |
|---|---|
| **Community Skills** | 100+ platform-curated Skills from publishers including Anthropic, OpenAI, and Genspark |
| **Team Skills** | Published by your organization, visible only to team members |
| **My Own Skills** | Skills you create, upload, or save yourself |
## Feature & How-To Guides {#feature-guide data-section-type=feature}
### Browsing and Using Community Skills {#community-skills}
The Skills library contains 100+ verified Skills. They are not organized by content category — instead, you filter across three dimensions:
| Filter | Options |
|---|---|
| **Publisher** | Anthropic, OpenAI, Genspark, and others |
| **Role** | Sales / Marketer / Product Manager / Researcher / Engineer / Designer / Operations / Founder / Creator / General |
| **Output type** | Document / Research / Data / Chart / Code / and more |
To use a Community Skill:
1. Go to **Skills** — the Community Skills tab is shown by default.
2. Use the Publisher, Role, and Output filters to narrow results.
3. Click a Skill card to view its description and preview.
4. Click **Add & Use** to install it.
5. Enter your prompt — AI generates output following that Skill's approach.
### Running a Skill {#running-a-skill}
1. **Select a Skill.** In the SkillBar, click **"Select a skill first"** to choose from your installed Skills, or type `/` to search by keyword. You can select multiple Skills at once.
2. **Enter your prompt.** Describe your request. Attach files if needed via the attachment icon or drag and drop.
3. **Watch execution in real time:**
| What you see | What it means |
|---|---|
| Execution step cards | Each operation (search, read file, write code) shown as a collapsible card |
| Thinking | The AI's reasoning process (expandable) |
| Streaming output | Real-time text output |
4. **Retrieve your files.** Deliverable files appear in the **Output files drawer** — click the folder icon in the top-right corner. Preview, download, or find past runs in the **Task List** (hamburger menu, top-left). Files are permanently stored in the cloud and can be downloaded at any time.
### Building Your Own Skill {#building-skills}
You don't need to be a developer. If you can describe how something should be done, you can create a Skill.
- **Create for Myself (AI-guided)** — Click **+ New Skill → Create for myself**. The skill-creator Agent guides you through name, description, inputs, outputs, and tool permissions. Test it, then **Save** — it appears in your My Own Skills tab.
- **Upload a Skill file** — Click **+ New Skill → Upload**. Drag and drop or select a file (`.zip`, `.skill`, or `.md`; max 200MB), then click **Save skill** — ready to use immediately. A `.zip` file must contain a `skill.md` with a name and description.
- **Create for Team** — the same flow as *Create for myself*, but the Skill is submitted for admin review. Once approved, it appears in every team member's **Team Skills** tab. Requires a Team Plan.
### Sharing Skills {#sharing-skills}
- **Peer Share (email invite)** — Open the Skill → click **Share** → enter the recipient's Genspark email. They accept via **"Shared with me"** on the Skills home page. Invites are valid for 30 days; recipients must have a Genspark account.
- **Team Publish** *(Team Plan required)* — Share → **Team** tab → **Publish to Team**. After admin approval (if required), the Skill appears in every member's Team Skills tab. When you update the Skill, members see an **"Update available"** prompt.
- **Public Link** — Share → **Public link** tab → **Create public link**. Copy and send the link — anyone can preview and install after signing in. Use **Regenerate** to invalidate the old link, or **Revoke** to disable it entirely. Public Link recipients install a snapshot copy that does not sync with future updates.
### Managing Your Skills {#managing-skills}
Click **Manage Skills** at the top of the Skills page to open the management panel.
| Tab | What it shows |
|---|---|
| **Installed** | All Skills you've installed |
| **Received** | Invitations shared with you via email |
| **Sent** | Skills you've shared with others |
- **Uninstall** — hover a Skill card → click Uninstall. Can be reinstalled at any time.
- **Update a Team Skill** — click the **"Update available"** badge when it appears.
- **Delete** — click Delete on any Skill in My Own Skills. This is permanent and cannot be undone.
- **Modify** — online editing is not supported. Edit locally → delete the old version → re-upload via **+ New Skill → Upload**.
### Connecting External Services (Connectors) {#connectors}
Connectors let Skills read live data from your external apps. Connect once; every Skill that needs the service uses it automatically.
To connect: click **Connectors** at the top of the Skills page, find the service using the search box, click **Install** (or **Connect**), and complete OAuth authorization. The service then shows as **"Connected."** To disconnect, click **Uninstall** or **Disconnect** next to any connected service.
| Category | Services |
|---|---|
| **Communication** | Gmail, Google Calendar, Outlook Email, Outlook Calendar, Microsoft Teams, Slack, LINE |
| **Productivity & Files** | Google Suite, Google Drive, Google Contacts, Outlook Contacts, Microsoft 365, OneDrive, SharePoint, Notion, Box, Confluence, DocuSign, Dropbox |
| **CRM & Sales** | Salesforce, HubSpot, Pipedrive, Mailchimp, Affinity CRM |
| **Developer** | GitHub, Jira, Linear, Deep Wiki, Chart Server |
| **Data Sources** | Hacker News, Supabase Database, Airtable |
| **Project & Ops** | Intuit QuickBooks, monday.com, ClickUp, Stripe, PayPal, Greenhouse, Ashby, Asana |
| **Other** | Google Chat, X (Twitter), Reddit, Merge.dev, Unified.to, Zapier, Ramp |
Some Skills require a Connector before running — they will block and prompt you to connect first.
### My Skills {#my-skills}
All Skills you create, import, or save appear under the **My Skills** tab — your personal collection, always one click away. From here you can launch, share, publish, or delete any Skill.
## Troubleshooting {#troubleshooting data-section-type=troubleshooting}
### A Skill ran but I can't find the output files — where do I look? {#find-output-files}
- Click the **folder icon** (top-right) to open the Output files drawer.
- Check the **Task List** (hamburger menu, top-left) for past run records.
- Review the execution step cards for any failed steps.
- If the Skill requires a Connector, confirm it's connected before re-running.
### My Peer Share invite expired — what do I do? {#invite-expired}
Open the Skill → Share → Email tab and resend. Invites expire after 30 days.
### I updated my Skill but team members don't see changes — why? {#update-not-showing}
Updates aren't pushed automatically. Members need to click the **"Update available"** prompt on their Skill card.
### I can't find the Team Publish option — why? {#no-team-publish}
Team Publish requires a Team Plan. Without one, clicking *Create for Team* or the Team tab redirects to the upgrade page.
## FAQ {#faq data-section-type=faq}
### What is the difference between a Skill and a template? {#vs-template}
A template reuses a layout. A Skill packages expert reasoning, content structure, and reusable prompts — and works across multiple Genspark products.
### When should I use a standard prompt instead of a Skill? {#vs-standard-prompt}
Use a prompt for one-off tasks. Use a Skill when you want the same structure, style, and reasoning applied consistently every time.
### When should I create a Skill instead of just prompting Super Agent? {#when-to-create}
When you're rewriting the same detailed instructions repeatedly, need consistent quality across a team, or want to share a proven workflow.
### Can I turn a good project into a Skill later? {#project-to-skill}
Yes — save any strong project as a Skill from My Skills. The approach is captured and immediately reusable.
### Do I need technical skills to create a Skill? {#need-technical-skills}
No. The *Create for myself* path is fully AI-guided — answer questions in plain language and the Agent builds it for you.
### Can I use Skills on mobile? {#mobile}
Skills run in the browser-based Super Agent Sandbox. Check the Genspark mobile app for the latest supported features.
### How do I share a Skill with my team without making it public? {#share-privately}
Use **Peer Share** for specific individuals, or **Team Publish** (Team Plan required) for your entire organization.
### Can I edit a Community Skill after adding it? {#edit-community-skill}
Not directly. Use it as a reference and create your own version with *Create for myself*.
# SecondBrain {#secondbrain data-doc-group=knowledge}
Source: https://www.genspark.ai/helpcenter/secondbrain
## What is SecondBrain? {#what-is-secondbrain data-section-type=overview}
**SecondBrain** is Genspark's personal knowledge and memory system — a unified hub that learns from your emails, meetings, files, and connected apps to give your AI a deep, personalized context about you and your work.
- **Connect your data** — with your permission, SecondBrain links to everything you already use: your email, calendar, chat history, meeting notes, external platforms, and more. It all becomes your personal context layer
- **Retrieve your SecondBrain data in Super Agent** — once connected, Super Agent has the full picture of who you are and how you work. This means smarter answers, more personalized actions, and a real edge whenever you need to pull together, organize, or make sense of information from multiple sources
---
## Quick Start {#quick-start data-section-type=overview}
- Open SecondBrain from its dedicated [website](https://www.genspark.ai/second-brain/home) or from the Genspark home page
- Connect at least one data source — your email, calendar, or an uploaded file is a good place to start
- Once connected, SecondBrain begins building your personal context layer automatically
- Head to the Genspark home page and open **Super Agent** — it will now draw on your SecondBrain data to give smarter, more personalized responses
---
## Feature & How-To Guides {#feature-guide data-section-type=feature}
### Step 1 — Collect Your Data {#step-1-collect-your-data}
SecondBrain pulls in your information from several sources:
- **Genspark projects** — auto-saves all your past Genspark activity, including AI Meeting Notes
- **Connected tools** — sync data automatically from all your sources
- **File upload** — manually upload documents, spreadsheets, or personal files directly into your memory up to 500 MB
### Step 2 — Connect Your Tools {#step-2-connect-your-tools}
To connect a data source:
- From the SecondBrain homepage, click **Connect** next to the tool you want to add
- Follow the authorization steps for that service
- Once connected, data begins syncing automatically in the background
Supported integrations: Gmail, Google Calendar, Outlook, Slack, Notion, HubSpot, Salesforce, Google Workspace, Microsoft 365, and more.
### Step 3 — Put Your SecondBrain to Work {#step-3-put-your-secondbrain-to-work}
Head to the Genspark homepage. From there, Super Agent can tap into all the memory, files, and data in your SecondBrain to get work done for you.
**Popular Plays** — common workflows that SecondBrain powers automatically once your sources are connected:
| Play | What it does | Sources used |
|---|---|---|
| Your weekly update, written for you | Generates a personalized weekly summary from your email, calendar, and connected apps | Gmail/Outlook, Calendar, connected apps |
| Your next meeting, briefed | Prepares a meeting brief with attendee context, relevant docs, and past conversations | Calendar, Docs, connected apps |
| The meeting's over — follow-ups already drafted | Turns your meeting notes into draft follow-up emails and action items | Meeting Notes, Gmail/Outlook |
| Never wonder "where did I put that?" | Searches across all your connected sources to find any file, email, or note | All connected sources |
Try prompts like:
- "Summarize my emails from today."
- "Draft an email and meeting invite for [name]."
- "Summarize last week's meetings into a weekly report."
- "Find the slides I made last week on [topic] and add a new page."
- "What to-dos did [partner] leave me in Slack today?"
- "Pull this quarter's numbers from my HubSpot and turn them into a deck."
- "Find everything in my Notion docs about [topic]."
## How Sync Works {#how-sync-works data-section-type=feature}
Your SecondBrain Note captures recordings on the device, then moves them to your phone and on to the web so you can read transcripts and summaries. Sync happens in two stages, and understanding them explains why a note sometimes appears on your phone before it shows up on the web.
### The two stages of sync {#the-two-stages-of-sync}
**Stage 1 — Device to phone.** The device records on its own and stores the recording locally. When it reconnects to your phone, it transfers the recording over Bluetooth (or over Wi-Fi if you use Fast Transfer). This transfer happens only when your phone is unlocked, Bluetooth is on, **and you've opened the Genspark app since the device reconnected.**
**Stage 2 — Phone to cloud to web.** Once the app has the recording, it uploads it, generates the transcript and summary, and syncs everything to the web (AI Meeting Notes).
The important part: the gate is on Stage 1. Recordings don't move off the device until the app has been opened.
### Opening the app is what starts the transfer {#opening-the-app-is-what-starts-the-transfer}
**Open the app, and keep it open until the transfer completes.** In the app, go to AI Notes — the note appears with a pending status and resumes automatically.
- Keep the phone unlocked with Bluetooth on while the transfer runs (the device-to-phone transfer is over Bluetooth).
- For a faster transfer, use **Fast Transfer** (the button in the top-right), which switches the connection from Bluetooth to Wi-Fi.
- Opening the app is the action that starts the transfer — it doesn't happen on its own while the app is closed.
## FAQ {#faq data-section-type=faq}
### Do I need the SecondBrain Note hardware to use SecondBrain? {#need-hardware}
No. SecondBrain is the software memory system — connect your apps and use in-app AI Meeting Notes with no device at all. The SecondBrain Note hardware is just one optional input channel for capturing in-person conversations.
### What's the difference between SecondBrain and SecondBrain Note? {#vs-secondbrain-note}
SecondBrain is the software: it links your data sources into a personal memory layer that Super Agent can draw on. SecondBrain Note is the physical recording device that feeds in-person meetings into that memory. You can use SecondBrain fully without the hardware. For more on SecondBrain Note, see [shop.genspark.ai](https://shop.genspark.ai/).
### How soon does my connected data show up in SecondBrain? {#data-sync-time}
Some integrations sync on a schedule rather than instantly, so allow a few minutes after connecting. If data is still missing after several minutes, disconnect and reconnect the source from the Home page.
### What data sources can I connect to SecondBrain? {#data-sources}
SecondBrain supports a growing list of connectors: email providers, calendars, messaging platforms, CRM tools, productivity suites, cloud storage, and manually uploaded files.
### Is my data private and secure? {#privacy-security}
SecondBrain may handle highly sensitive personal context — your emails, meetings, files, and conversations — so security and privacy are core product considerations. Genspark maintains an enterprise-grade security framework and is certified with **SOC 2 Type II** and **ISO 27001**. Your SecondBrain data is only accessible to you. Connected integrations require your explicit authorization and can be disconnected at any time from the Home page.
# SecondBrain Note {#secondbrain-note data-doc-group=knowledge}
Source: https://www.genspark.ai/helpcenter/secondbrain-note
## What is SecondBrain Note? {#what-is-secondbrain-note data-section-type=overview}
SecondBrain Note is Genspark's hardware device that captures, transcribes, and summarizes everything you hear and say, turning real-world conversations into structured, searchable notes automatically.
- **Capture from the real world** — Record meetings, conversations, and ideas using dedicated hardware or the Genspark mobile app
- **AI-powered output** — Transcriptions and full summaries generated automatically, no manual note-taking required
- **Deeply integrated** — Syncs with the Genspark app, GenMail, and your SecondBrain data sources for a unified AI memory
SecondBrain Note can be purchased at [shop.genspark.ai](https://shop.genspark.ai/).
## Quick Start {#quick-start data-section-type=overview}
### Step 1: Download the Genspark App {#step-1-download-the-genspark-app}
Download "Genspark" in the App Store or Google Play and complete user registration.
### Step 2: Device Pairing {#step-2-device-pairing}
1. Press and hold the device button for 3 seconds to power on — it enters pairing mode automatically and the blue indicator light flashes.
2. Open the Genspark app and tap **AI Notes** at the bottom.
3. Tap **Start Pairing**:
- Tap the **SecondBrain Note** button from the homepage, or
- Tap **AI Notes** and then **+ Device**.
4. Match the last 4 digits of the serial number (SN) on the packaging to identify your device.
5. Tap your device to connect — the blue light turns off when paired.
6. Tap **Get Started** to begin.
**Note:** If connecting in a room with many devices, use the SN search box to find yours quickly. A scrollable list appears when multiple devices are detected.
### Using the MagSafe Wallet {#using-the-magsafe-wallet}
- Place the SecondBrain Note into the inner pocket of the wallet.
- With the cardholder opening facing down, attach magnetically to the back of your phone.
- If your phone doesn't support magnetic attachment, use the included magnetic strip on your phone case.
- The wallet doubles as a phone stand.
**Phone case compatibility:** The device works through most standard phone cases for general recording. For phone call recording specifically, thinner cases work best since the V.C.S. sensor picks up vibrations through the phone's speaker. Very thick rugged cases may reduce call recording quality but won't affect open-air meeting or voice memo recording. If call recordings sound muffled, try a thinner case or remove the case during calls.
### Charging the Device {#charging-the-device}
SecondBrain Note uses magnetic charging.
1. Connect the included USB-C cable (Type-C input, 4-pin magnetic output) to the device.
2. On successful connection: indicator light flashes green + single vibration.
3. Fully charged: indicator light stays solid green.
4. Disconnect the cable — light turns off.
Charging works the same way with the device inside the MagSafe wallet.
## What's in the Box {#whats-in-the-box data-section-type=feature}
- SecondBrain Note × 1
- MagSafe Wallet × 1
- USB-C Charging Cable × 1
- Metal Ring × 1
- Quick Guide × 1
- Safety & Warranty Guide
## Recording Consent & Privacy {#recording-consent-and-privacy data-section-type=feature}
Before recording other people, make sure you comply with applicable recording, privacy, and biometric laws. Laws vary by country and by U.S. state, and some jurisdictions require consent from all participants before a conversation can be recorded. If participants are in different locations, follow the stricter rule.
You are responsible for:
- Notifying participants before recording starts
- Obtaining any consent required by law
- Stopping the recording if someone does not consent
- Avoiding recording in sensitive or confidential settings unless you are clearly permitted to do so
SecondBrain does not obtain consent on your behalf or determine whether your recording is lawful. If you are unsure, consult your legal counsel.
## Recording Notes {#recording-notes data-section-type=feature}
- **Starting a recording:**
- **Hardware button** — Long-press for 2 seconds (white breathing light appears, single vibration)
- **In-app** — Tap the new button inside AI Notes and choose "Record with SecondBrain Note"
- **Marking highlights:** During a recording, a short press on the device button marks an important moment (a single vibration). These appear as highlights in your generated notes for quick reference.
- **Ending a recording:** Long-press the button for 2 seconds — LED turns off, two vibrations confirm.
Output depends on recording length:
| Recording Length | Output |
| --- | --- |
| Under 2 minutes | Transcription only |
| Over 2 minutes | Full summary + transcription |
Real-time transcription is not available in the current release. Notes are generated after the recording ends.
**Note Source Indicator:** A device icon appears next to any note created via SecondBrain hardware, distinguishing it from notes made through other methods.
## Data Transfer {#data-transfer data-section-type=feature}
After recording, audio files save automatically on the device. To transfer:
1. Connect the device to the Genspark app via Bluetooth.
2. Files transfer automatically.
**Quick Transfer (Wi-Fi):**
- After connecting, a "Fast Transfer" prompt appears in the top-right corner — tap it.
- The app switches from Bluetooth to a direct Wi-Fi connection with the device.
- Stay in the app until the transfer completes.
- The app automatically restores Bluetooth mode when finished.
## Offline Usage {#offline-usage data-section-type=feature}
**Can SecondBrain be used without a phone?**
Yes — the device records independently while your phone is off. Once your phone reconnects, notes transfer automatically and summaries are generated. However, the device requires a phone connection for full functionality (transcription, summaries, syncing).
## Adding Notes {#adding-notes data-section-type=feature}
Tap **Add New** in the app to create notes in three ways:
| Method | Description |
| --- | --- |
| AI Meeting Record With Phone | Uses the app's built-in AI (no hardware required) |
| Record with SecondBrain | Uses your connected SecondBrain hardware |
| Import Audio File | Upload an existing audio file for transcription |
**Photo Integration:** During a recording, take a photo directly from the app — it embeds in your generated notes with optional accompanying text.
## SecondBrain Data Sources {#secondbrain-data-sources data-section-type=feature}
You can connect SecondBrain to a variety of different sources, including first-party data, imported data, external connectors, or manual entry. For example, email, CRM applications, Genspark projects, and more.
## Super Agent Integration {#super-agent-integration data-section-type=feature}
Once your data sources are connected, the Super Agent can act on your SecondBrain data:
- Check and triage your email
- Query records in HubSpot or Salesforce
- Create slides from your notes
- Summarize meeting recordings
- Generate weekly reports
## Credits & Billing {#credits-and-billing data-section-type=feature}
### How It Works {#how-it-works}
SecondBrain Note separates recording from processing:
- Recording happens entirely on the device — no internet, no credits, no limits. Press the button, talk, done.
- Processing (transcription + AI summary) happens after the recording syncs to Genspark. This is where minutes and credits come in.
When a recording syncs, the app automatically transcribes the audio and generates a summary. That processing step is what counts against your monthly minutes.
### What's Free {#whats-free}
Every SecondBrain Note purchase includes 300 free transcription minutes per month. No Genspark subscription required. These minutes cover both transcribing your recordings and generating AI summaries.
The 300-minute allowance resets each month.
### After Your Free Minutes {#after-your-free-minutes}
| Membership | Monthly Transcription | After Free Minutes |
| --- | --- | --- |
| No subscription (Note hardware only) | 300 minutes/month included | ~5 credits per minute |
| Plus | Unlimited | — |
| Pro | Unlimited | — |
**Without a subscription:** Once you've used your 300 included minutes for the month, new recordings that sync draw from your daily free credit allotment (100 credits/day) at approximately 5 credits per minute. When your daily credits run out, processing pauses until they reset the next day. You can also purchase credits individually or sign up for a Plus or Pro plan for unlimited transcription.
**With Plus or Pro:** Transcription and summarization are unlimited — record as much as you want without worrying about minutes or credit costs. This applies across all your devices.
## Warranty & Returns {#warranty-and-returns data-section-type=feature}
For full details, see the Safety & Warranty Guide. To file a warranty claim, contact support@genspark.ai with a description of the issue and photos or video showing the defect.
**Returns:**
- Returns are accepted within 30 days of receipt.
- Only unopened, sealed boxes are eligible for return.
- No returns accepted after the 30-day window.
- Device must be in original condition with all packaging and accessories.
- If your device is defective or arrived damaged, it's covered under warranty (see above).
**Who pays for return shipping:**
- Unopened returns — you are responsible for the cost of the return shipping label.
- Faulty or defective product — Genspark covers the return shipping cost.
## Recording Consent, Privacy & Responsible Use {#recording-consent-privacy-and-responsible-use data-section-type=feature}
Recording laws vary by country and by U.S. state. In some places, you may need consent from all participants before recording. Because rules differ, you should always be transparent, notify people before recording starts, and obtain any consent required by law.
SecondBrain does not obtain consent on your behalf or determine whether a recording is lawful.
### Before You Record {#before-you-record}
Before starting a recording, make sure you:
- Tell people clearly that you want to record
- Obtain any consent required by law
- Confirm that the setting is appropriate for recording
- Avoid recording private, sensitive, or confidential conversations unless you are clearly permitted to do so
- Follow the stricter rule if participants are in different locations
A device light, app status, or other visible indicator may help signal that recording is happening, but that alone is not enough. You should still notify people directly.
### Simple Ways to Notify People {#simple-ways-to-notify-people}
You can use plain language like:
- "I'd like to record this so I can review my notes later. Is that okay with everyone?"
- "Just a heads-up: I'm using SecondBrain to record and summarize this conversation."
- "Before we continue, I want to let you know this conversation may be recorded. Please let me know if you're not comfortable with that."
### When Consent May Be Required {#when-consent-may-be-required}
Consent requirements depend on where people are located and what kind of conversation is being recorded.
Keep these points in mind:
- Recording laws vary by country and by U.S. state
- Some jurisdictions require consent from all participants before recording
- If participants are in different locations, you should follow the stricter rule
- You are responsible for understanding whether consent is required in your situation
### Best Practices {#best-practices}
Use these best practices every time you record:
- Notify participants before recording starts
- Be specific about what is being recorded
- Ask before beginning, not after
- Give people a real chance to decline
- Stop recording when the conversation ends
- Record only what you need
- Handle recordings responsibly and share them only as appropriate
**Good practice:**
- Explain what SecondBrain is doing
- Ask for permission in a clear, direct way
- Make sure people understand they are being recorded
**Avoid:**
- Starting a recording without notice
- Relying only on a device indicator or screen status
- Leaving recording on longer than needed
- Recording "just in case" in situations that may be sensitive
### Sensitive Situations {#sensitive-situations}
You should be especially careful in sensitive or confidential settings.
Avoid recording unless you are clearly permitted to do so, including situations such as:
- Healthcare or medical conversations
- Legal or financial discussions
- HR or workplace investigations
- School or student-related situations
- Private customer, client, or patient information
- Conversations involving confidential business information
- Any place where people reasonably expect a high level of privacy
When in doubt, do not record.
### What to Do If Someone Says No {#what-to-do-if-someone-says-no}
If someone does not consent:
- Do not start recording.
- Stop recording immediately if it has already started.
- Do not pressure the person to agree.
- Continue without recording, if appropriate.
- Delete the recording if it captured someone who did not agree and you are not clearly permitted to keep it.
Respecting someone's choice is part of responsible use.
## Need More Help? {#need-more-help data-section-type=feature}
For additional device guides and troubleshooting, visit the SecondBrain support page. For anything not covered there, email support@genspark.ai.
## Troubleshooting {#troubleshooting data-section-type=troubleshooting}
### Device won't appear when scanning for nearby devices {#device-wont-appear-when-scanning-for-nearby-devices}
- Make sure Bluetooth is enabled on your phone.
- Keep the hardware clip within close range of your phone during pairing.
- Check that the clip has sufficient battery charge.
- Restart the clip and try scanning again.
- If multiple devices appear, use the SN search box to locate yours by the last 4 digits of the serial number on the box.
### Recording started or stopped unexpectedly {#recording-started-or-stopped-unexpectedly}
- Avoid placing the clip where it may be accidentally pressed.
- Check the clip's button sensitivity — a short press marks a highlight; a long press starts or stops recording.
- If an accidental recording was created, you can delete it from the notes list in the app.
### Notes didn't sync after the meeting {#notes-didnt-sync-after-the-meeting}
- Make sure your phone is unlocked and the Genspark app has been opened at least once since reconnecting.
- Check that Bluetooth is still active — the clip syncs over Bluetooth, not Wi-Fi.
- If the sync still hasn't completed, open the app and navigate to AI Notes — the note should appear with a pending status and resume automatically.
- If the note is missing entirely, check that the clip had enough storage and battery to complete the recording.
### Summary wasn't generated — only a transcript appeared {#summary-wasnt-generated-only-a-transcript-appeared}
- Recordings under 2 minutes produce a transcript only. If your meeting was longer, check that the full recording synced correctly.
- If the recording was interrupted mid-session, only the synced portion will be processed.
### App lost connection to a data source (Gmail, Outlook, Notion, etc.) {#app-lost-connection-to-a-data-source}
- Integration tokens can expire over time, especially for Outlook.
- Go to Home → select your account, disconnect the affected account, and reconnect it.
- After reconnecting, your data will resume syncing automatically.
### SecondBrain isn't reflecting recent emails or files {#secondbrain-isnt-reflecting-recent-emails-or-files}
- Some integrations sync on a schedule rather than instantly — allow a few minutes after connecting.
- If data is still missing after several minutes, disconnect and reconnect the integration from Home → Feed Your Brain.
## FAQ {#faq data-section-type=faq}
### Is my data secure? {#is-my-data-secure}
Your recordings are encrypted during transfer and processed through Genspark's enterprise-grade security framework, which is certified with SOC 2 Type II and ISO 27001. Once a recording is transferred to the app, the local file on the device is deleted immediately. Your data is only accessible to you.
### How do I use the device across multiple platforms? {#how-do-i-use-the-device-across-multiple-platforms}
On a PC, download the Genspark software or log in via the web browser at genspark.ai. After logging in with your account, you can access shared cloud recording files and historical information across multiple devices.
### How do I update my device firmware? {#how-do-i-update-my-device-firmware}
Before updating, make sure: (1) the device has sufficient battery power, (2) the device is not in recording mode, and (3) the device is connected to the Genspark app. Then follow the in-app prompts to install the update. Keeping firmware up to date ensures the best and most current user experience.
### Can it be shared by multiple users? {#can-it-be-shared-by-multiple-users}
The device cannot connect to two phones at the same time. Once connected to one phone, other phones won't be able to discover or connect to it. However, you can log in with the same account on another phone to connect to the device and access cloud data. You can also view history on the PC or web version. To switch to a different account, unpair the device from the current account first.
### How do I find my serial number (SN)? {#how-do-i-find-my-serial-number-sn}
You can find it in two places: (1) on the label printed on the recorder's packaging box, or (2) in the Genspark app — connect your device and go to the device settings page where the serial number is displayed.
### Can I use headphones when recording a phone call? {#can-i-use-headphones-when-recording-a-phone-call}
No. In call recording mode, the vibration conduction sensor (V.C.S.) captures voices through vibration from the phone's earpiece speaker. Using a headset while recording calls is not supported.
### How do I force a restart? {#how-do-i-force-a-restart}
If the device freezes or becomes unresponsive, press and hold the button until the device reboots with a single vibration. This will not clear any pairing or connection information.
### What recording modes does the device support? {#what-recording-modes-does-the-device-support}
Two modes: (1) Open recording — for meetings, voice memos, interviews, lectures, medical use cases, etc. (2) Phone call recording — place the device in the magnetic card holder and attach it to the back of your phone. The vibration conduction sensor picks up sound directly from the phone's speaker.
### Does it support continuous recording? {#does-it-support-continuous-recording}
Yes. With a full charge, it supports up to 35 hours of continuous recording.
### How many minutes are free, and what happens after that? {#how-many-minutes-are-free-and-what-happens-after-that}
**For free users:** The SecondBrain Note hardware includes 300 free minutes each month. Once you've used those, transcription and summaries draw from your daily free credit allotment (100 credits/day, ~5 credits per minute). When your daily credits run out, processing pauses until they reset the next day — or you can upgrade to a Plus or Pro plan for unlimited transcription. Recording itself is always free; only AI processing uses credits.
**For Paid members (Individual, Team, and Enterprise):** You can get unlimited meeting notes across all channels — up to 24 hours per day, tracked by member ID and shared across your devices.
### What should I do if my device is lost? {#what-should-i-do-if-my-device-is-lost}
Don't worry about data leakage — after each recording is transferred to the app, the local file on the device is deleted immediately. If the device has been linked to your account, others won't be able to access it or bind it to another account. You can still log in to the app to access your recordings.
### Where is the microphone located? {#where-is-the-microphone-located}
The device has four MEMS microphones plus one vibration conduction sensor (V.C.S.). The four MEMS microphones are at the top and on the left and right sides of the card. The V.C.S. is located inside the device.
### Can I connect multiple SecondBrain Note devices to one account? {#can-i-connect-multiple-secondbrain-note-devices-to-one-account}
### My device is defective or arrived damaged. What do I do? {#my-device-is-defective-or-arrived-damaged-what-do-i-do}
Email support@genspark.ai with a description of the issue and photos or video of the defect. The team will review your case and, if confirmed as a product quality issue, send a replacement unit at no charge. You'll be responsible for shipping the defective unit back (see return addresses above).
### Where does SecondBrain Note deliver? {#where-does-secondbrain-note-deliver}
SecondBrain Note currently ships to United States, Japan, and South Korea. We are actively working to expand to more regions. If you are outside these countries, check [shop.genspark.ai](https://shop.genspark.ai/) for latest availability.
### Need a replacement cable or additional accessories? {#need-a-replacement-cable-or-additional-accessories}
The charging cable and accessories are not available to purchase on its own. If your cable is defective, lost, or worn out, email support@genspark.ai.
# GenMail {#genmail data-doc-group=office-suite}
Source: https://www.genspark.ai/helpcenter/genmail
## What is GenMail? {#what-is-genmail data-section-type=overview}
**GenMail** is Genspark's AI-powered email and calendar client. It connects your Gmail and Outlook accounts into a single, intelligent workspace to help you read, write, and manage emails faster with AI that learns how you work.
GenMail can:
- **Triage** your inbox in minutes
- **Draft replies** in your voice
- **Manage** multiple accounts
- **Schedule** sends
- **Search** email with natural language
- **View** a unified calendar
GenMail is organized into four main views:
| Tab | What it does |
| --- | --- |
| **Super Agent** | AI inbox triage — categorizes emails, generates summaries, suggests follow-up actions |
| **Inbox** | Traditional email view — full inbox, folders, and search |
| **Calendar** | Unified calendar combining Google and Outlook events with multi-timezone support |
| **Email Brain** | Your personal AI profile — see what GenMail has learned about you and correct anything |
---
## Quick Start {#quick-start data-section-type=overview}
### Prerequisites {#prerequisites}
- A Genspark account
- A Gmail and/or Outlook email account
### Set up in 5 minutes {#setup-in-5-minutes}
1. **Download GenMail** — available on Mac, Windows, iOS, and Android
2. **Sign in** with your Genspark account
3. **Connect your email** — add your Gmail and/or Outlook accounts. GenMail begins syncing your inbox and calendar automatically.
4. **Email Brain analyzes your email history** — GenMail analyzes your email history in the background to build your profile: writing voice, frequent contacts, active projects, and communication patterns
5. **Start using it** — your inbox will be organized and ready to go
| Platform | Availability |
| --- | --- |
| **Mac** | Desktop client |
| **Windows** | Desktop client |
| **iOS** | Mobile app |
| **Android** | Mobile app |
| Email Provider | Status |
| --- | --- |
| **Gmail** | Fully supported (personal and workspace) |
| **Outlook** | Fully supported (personal, enterprise, and shared mailboxes) |
> **Standalone client:** GenMail connects directly to your email provider. Your emails stay in Gmail/Outlook, and you can use GenMail alongside your existing email app or switch entirely.
---
## Feature & How-To Guides {#feature-guide data-section-type=feature}
### Super Agent — Inbox Triage {#super-agent-inbox-triage}
The Super Agent automatically categorizes your emails by importance so you can skip the scroll and focus on what matters.
| Category | What it means |
| --- | --- |
| **Needs You** | Emails requiring your action — urgent, time-sensitive, or directly addressed to you |
| **Upcoming Today** | Calendar invites and scheduled events (meetings, calls, etc.) |
| **FYI** | Worth knowing about, but no response needed |
| **Other** | Newsletters, ads, subscriptions, and other low-priority messages |
Each email is **summarized** so you can understand it without opening it. The Super Agent also flags attachments, suggests follow-up actions (like drafting a reply or reading an attachment), and maintains a task history.
### AI Write — Draft Emails with GenMail {#ai-write}
GenMail drafts replies that match your personal writing style, learned by Email Brain. You stay in control of every message — GenMail learns from edits you make to drafts, and the more you use it, the more accurate your voice match becomes.
| Action | How it works |
| --- | --- |
| **Generate a draft** | AI writes a first pass based on email context and your voice |
| **Adjust tone** | Ask it to be "more professional," "shorter," "friendlier," etc. |
| **Give specific instructions** | e.g., "acknowledge receipt and ask for the deadline" |
You review and approve AI-drafted replies before they send.
### Skills — Pre-Built AI Actions {#skills}
| Skill | What it does |
| --- | --- |
| **Morning Brief** | Summarizes your inbox for the day |
| **Catch Me Up** | Summarizes what you missed (after time away, weekends, etc.) |
| **Organize Inbox** | Sorts and categorizes your messages |
| **Find It For Me** | Searches your email history using natural language |
**Example — Morning Brief:** gives you a daily triage — a quick summary of what came in overnight, what needs attention, and what you can safely skip. It's designed to be the first thing you check so you can prioritize your day in minutes instead of scrolling through dozens of emails.
> **Note:** Skills consume credits but can be turned off.
### Compose: Chinese, Japanese, and Korean fonts {#compose-chinese-japanese-and-korean-fonts}
Compose now includes Chinese, Japanese, and Korean typefaces available from the compose font picker — so you can set the typeface when you write multilingual drafts, instead of the fallback rendering.
### Mark all as read {#mark-all-as-read}
Right-click the inbox and choose Mark all as read — it clears unread mail in one action across all your accounts, no jumping between mailboxes to clear unread counts one by one.
### Preview, print, and save as PDF {#preview-print-and-save-as-pdf}
You can now preview an email, print it, or save it as a PDF — useful for archiving, offline reading, or attaching a copy elsewhere.
### Calendar: share colleague calendars {#calendar-share-colleague-calendars}
You can view a colleague's calendar and share yours with them, so the team can see each other's availability directly in GenMail.
The calendar now shows a free/busy panel — when you're scheduling, check when your colleagues are free before you pick a time, instead of guessing or emailing back and forth.
### Schedule Send {#schedule-send}
Write an email now, send it later. Pick a specific date and time for delivery. Scheduled emails appear in a dedicated **Scheduled** folder where you can review or cancel them before they go out.
### Undo Send {#undo-send}
When you send a normal email, a short countdown starts and the message doesn't actually go out until the countdown ends — so you can cancel it before it leaves. Applies automatically to normal sends; cancel directly from the confirmation during the countdown (no Scheduled folder).
**How it differs from Schedule Send:** Schedule Send = send at a chosen future time, cancel from the Scheduled folder. Undo Send = automatic short countdown on every normal send, cancelled in place. Use Schedule Send to time an email; Undo Send is your safety net on everything else.
### GenTeam Integration {#genteam-integration}
- **Email an agent directly** — send an email to an agent and it can act on it
- **@-mention agents in a thread** — bring an agent into an email thread
- **Deliverable cards with in-app previews** — slides and docs appear as cards you can preview inline
### Calendar {#calendar}
GenMail includes a unified calendar that combines events from your connected Gmail and Outlook accounts.
| Feature | Details |
| --- | --- |
| **Create events** | Click and drag (desktop) or long-press (mobile) |
| **RSVP tracking** | See attendee responses at a glance |
| **AI booking** | Super Agent can automatically book events to your calendar |
### Email Signatures {#email-signatures}
GenMail supports custom email signatures. Create a new signature or import your existing one from Gmail or Outlook.
### The Email Brain — Your AI Profile {#email-brain}
The Email Brain tab shows everything GenMail has learned about you:
- Your identity and workplace
- Your writing voice and reply style
- Your frequent contacts and relationship context
- Your active projects and conversations
**If it gets something wrong:** click the **Fix** button to correct any inference. The AI updates its memory and adjusts future behavior accordingly. The more you correct, the better it gets.
### Privacy and Security {#privacy-and-security}
GenMail connects to your Gmail and Outlook accounts using **OAuth 2.0**. You sign in through your email provider's own login screen. GenMail never sees or stores your email password.
| Provider | Authorization Method |
| --- | --- |
| Gmail | Google OAuth 2.0 |
| Outlook | Microsoft OAuth 2.0 |
You can revoke GenMail's access at any time from your Google or Microsoft account settings without affecting your email or Genspark account.
**Where your emails live:** GenMail is a client, not a storage layer. Your emails stay in Gmail and Outlook — GenMail reads and writes through their APIs but does not maintain a separate copy of your mailbox on its own servers.
| What stays in your email provider | What GenMail stores |
| --- | --- |
| All emails, attachments, folders | Your Email Brain profile (writing style, contacts, patterns) |
| Sent mail history | Contact list |
| | Scheduled send emails |
| | Calendar events |
### Notifications {#notifications}
GenMail sends a system notification when new mail arrives, so you can stay on top of your inbox without keeping the app open. This works on both desktop (Mac / Windows) and mobile (iOS / Android).
> **Tip:** Make sure GenMail has notification permissions enabled at the system level:
>
> - iOS: **Settings → GenMail → Notifications**
> - Android: **Settings → Apps → GenMail → Notifications**
> - Mac: **System Settings → Notifications → GenMail**
> - Windows: **Settings → System → Notifications → GenMail**
### Settings {#settings}
**Account & Connections**
To add an email account:
1. Go to **Settings → Connected Accounts**
2. Click **Add mail account**
3. Sign in through your email provider's login screen
4. GenMail begins syncing your inbox and calendar automatically
To remove an email account:
1. Go to **Settings → Mail Accounts**
2. Click **Remove** next to the account
3. GenMail stops syncing — your emails in Gmail/Outlook are unaffected
To change your default sending account:
1. Go to **Settings → Mail Accounts**
2. Drag and reorder your accounts — the top account is your primary
**Email Preferences**
To create or update your email signature:
1. Go to **Settings → Email Preferences → Signature**
2. Write a new signature or click **Import** to pull in your existing Gmail or Outlook signature
3. If you have multiple accounts connected, set a signature for each one individually
To schedule an email:
1. Compose your email as normal
2. Click the dropdown arrow next to **Send**
3. Select **Schedule Send** and pick your date and time
4. The email appears in your **Scheduled** folder — open it anytime to edit or cancel before it sends
**AI & Email Brain**
To review what the Email Brain has learned:
1. Open the **Email Brain** tab from the main navigation
2. Browse your inferred profile — writing voice, frequent contacts, active projects
To correct an inference:
1. Find the item you want to fix in the **Email Brain** tab
2. Click **Fix**
3. Enter the correct information — the AI updates immediately and adjusts future behavior
To change your AI draft tone:
1. Go to **Settings → AI & Email Brain → AI Draft Style**
2. Choose **Professional**, **Balanced**, or **Casual**
3. You can still override tone per email by giving GenMail instructions in the compose window
To turn a Skill on or off:
1. Go to **Settings → AI & Email Brain → Skills**
2. Toggle any Skill (Morning Brief, Catch Me Up, etc.) on or off — active Skills consume credits, disabling unused ones preserves your balance
---
## FAQ {#faq data-section-type=faq}
### What email providers are currently supported? {#what-email-providers-are-currently-supported}
Right now only Gmail and Outlook are supported.
### How do I view my credit history? {#how-do-i-view-my-credit-history}
Visit the Genspark credit usage page via **Settings**.
### Can I see colleague availability in the calendar? {#colleague-availability-calendar}
Yes. The calendar's free/busy panel shows when your colleagues are free when you're scheduling, and you can view a colleague's calendar or share yours with them.
### Does GenMail store my emails on Genspark's servers? {#genmail-store-emails-on-servers}
No. GenMail connects directly to your email provider and your emails stay there — GenMail is not a separate storage layer.
### Can I turn Skills off? {#turn-skills-off}
Yes, they can be turned off in **Settings**.
### Will GenMail send emails without my approval? {#send-without-approval}
No, replies are always shown for your review and approval before sending.
### Can I connect more than two email accounts? {#connect-more-than-two-accounts}
Yes.
### What happens if I correct something in the Email Brain? {#correct-email-brain}
Clicking **Fix** updates the memory for that inference. It adjusts future behavior immediately — the more you correct, the more personalized it becomes.
### Can I print an email or save it as a PDF? {#print-save-as-pdf}
Yes. Open the email and use the print / save-as-PDF option — useful for archiving or sharing a copy outside GenMail.
### How do I mark everything in my inbox as read? {#mark-inbox-as-read}
Right-click the inbox and choose Mark all as read — it clears unread mail in one action across all your connected accounts.
### How do I add a contact to my CRM from an email? {#add-contact-to-crm}
Use the CRM this contact skill on the email view — it saves the contact into your CRM automatically.
### Does my calendar work with my team's? {#calendar-works-with-team}
Yes. Colleague calendars can be shared and viewed in GenMail, and the free/busy panel shows availability when you schedule.
# GenClipboard {#genclipboard data-doc-group=tools}
Source: https://www.genspark.ai/helpcenter/genclipboard
## What is GenClipboard? {#what-is-genclipboard data-section-type=overview}
GenClipboard is Genspark's free native clipboard manager for Mac, Windows, and Linux. It remembers everything you copy and gives you a searchable local history — all keyboard-first, so you never lose a snippet or reach for the mouse.
### Core capabilities {#core-capabilities}
- **Clipboard history** — captures every copy (text, rich text, images, files, URLs, JSON), fully searchable, stored locally, with automatic deduplication
- **Instant recall** — one hotkey opens the panel; type to search and press 1–9 to paste
- **Pinned clips** — pin the snippets you use often so they never scroll away
- **Cross-device sync** — optional and end-to-end encrypted; text and rich-text clips sync across your devices, while images and files always stay local
- **Screenshot & annotate** — capture a region, mark it up with arrows, shapes, text, and mosaic, then copy the result back to your clipboard
- **Scrolling capture** — capture content taller than the screen as one long image
- **Screen recording** — capture a region as an MP4
- **In-app translation** — highlight text in any app and replace it in place or copy the translation. Powered by your Genspark account
- **Image translation (Mac & Windows)** — recognize text inside any image clip with on-device OCR, then translate it through your Genspark account
## Quick Start {#quick-start data-section-type=overview}
1. **Download and install GenClipboard.** Download from [genclipboard.ai](https://genclipboard.ai) — choose Mac, Windows, or Linux. On Mac, drag GenClipboard to your Applications folder; on Windows, run the installer.
2. **Grant permissions (Mac only).** Go to System Settings → Privacy & Security → Accessibility and enable GenClipboard. The first time you take a screenshot or recording, macOS also asks for Screen Recording permission. Windows and Linux need no extra permission.
3. **Open the panel.** Press `⇧⌘V` (Mac) or `Ctrl+Shift+V` (Windows / Linux). The panel appears at the bottom of your screen with everything you've copied.
4. **Find and paste.** Type to search, use the arrow keys to navigate, press 1–9 or Enter to paste a clip, and `Esc` to dismiss.
## Keyboard Shortcuts {#keyboard-shortcuts data-section-type=feature}
GenClipboard is built to run entirely from the keyboard. Defaults:
| Shortcut (Mac / Windows & Linux) | Action |
|---|---|
| `⇧⌘V` / `Ctrl+Shift+V` | Toggle the clipboard panel |
| `1–9` | Quick-paste the Nth clip |
| `↵ Enter` | Paste the selected clip |
| `← →` | Navigate between clips |
| `⌘P` / `Ctrl+P` | Pin / unpin a clip |
| `⇧⌘A` / `Ctrl+Shift+A` | Screenshot & annotate |
| `⇧⌘S` / `Ctrl+Shift+S` | Scrolling capture |
| `⇧⌘R` / `Ctrl+Shift+R` | Screen recording |
| `Esc` | Dismiss the panel |
| `?` | Show all shortcuts |
Every hotkey above is configurable in Settings. The translation hotkeys ship unassigned — set them up in Settings → Translation (see the FAQ below).
## Privacy {#privacy data-section-type=feature}
Your clipboard stays yours.
- **Local-first** — your clip history is stored locally in a database on your own device
- **Encrypted sync** — sync is opt-in and end-to-end encrypted. Your encryption password never leaves your device, so the cloud can never read your clip content
- **Password managers are ignored** — copies from 1Password, Bitwarden, KeePassXC, and other password managers are never recorded by default
- **No content telemetry** — usage analytics track app events only and are designed to never include the content of what you copy
## FAQ {#faq data-section-type=faq}
### What does GenClipboard cost? {#what-does-genclipboard-cost}
It's free — no subscription, no credits. Translation is also free and only requires signing in with your Genspark account.
### Does it sync between devices? {#does-it-sync-between-devices}
Yes, optionally. Sync is end-to-end encrypted and merge-aware. Text and rich-text clips sync across your devices; images and files always stay local. Set it up in Settings: sign in with your Genspark account, then create a sync encryption password, it stays on your devices and is never sent to the server.
### How many clips does it keep? {#how-many-clips-does-it-keep}
By default GenClipboard keeps your 500 most recent clips locally, deleting the oldest unpinned ones first, pinned clips are never removed. You can change the limit in Settings → Max Clipboard History.
### Is my clipboard data private? {#is-my-clipboard-data-private}
Yes. GenClipboard is local-first, sync is end-to-end encrypted, and there's no telemetry on your clipboard content.
### How do I use translation? {#how-do-i-use-translation}
Open Settings → Translation, sign in with your Genspark account, and assign a hotkey to one or both modes: Selection Translate (a popup appears next to your highlighted text, with Replace and Copy buttons) and Silent Translate (the translation replaces your highlighted text in place, no popup). Set your First and Second Languages there too, GenClipboard translates into your First Language, or into your Second Language when the source is already in your First.
### Can I change the hotkey? {#can-i-change-the-hotkey}
Yes — set your own panel hotkey in Settings → Toggle Hotkey. The screenshot, scrolling-capture, recording, and translation hotkeys are configurable in Settings as well.
### How does GenClipboard fit with GenTerminal? {#how-does-genclipboard-fit-with-genterminal}
Both are part of Genspark's native desktop tools — GenClipboard manages your clipboard, GenTerminal is your SSH terminal. They're independent apps that share your Genspark account.
### How do I send diagnostic information when I report a problem? {#how-do-i-send-diagnostic-information-when-i-report-a-problem}
If you hit an issue and our support team asks for logs, open Settings and click Send diagnostics. This uploads recent app logs so we can investigate — your clipboard content is not included. Mention the approximate time the problem happened when you contact support.
# Connecting Salesforce to Genspark {#connecting-salesforce-to-genspark data-doc-group=tools}
Source: https://www.genspark.ai/helpcenter/salesforce-integration
## What is the Salesforce Integration? {#what-is-the-salesforce-integration data-section-type=overview}
Connect your Salesforce account to Genspark once, and every Genspark surface can use it — Claw, Super Agent, Workflow Builder, and AgentBase all share the same connection. Authorize it a single time, then pull records, update data, run reports, and automate CRM work across Genspark.
> **Before you start:** Claw requires a Genspark Cloud Computer subscription (or the local desktop app) plus active credits. Super Agent, Workflow Builder, and AgentBase work with a standard Genspark account.
## Connect Salesforce {#connect-salesforce-one-time-setup data-section-type=feature}
Genspark connects to Salesforce through its secure OAuth connector — you authorize Salesforce once from your Genspark account, in your normal browser, and the connection is saved for every surface.
- **Step 1 — Start the connection:** In Genspark, go to **Skills → Connectors → CRM & Sales**, find **Salesforce**, and click **Connect**.
- **Step 2 — Complete the OAuth login:** Sign in to Salesforce through the standard OAuth login that opens in your browser. Approve access and you're done.
That's it — you won't need to reconnect for each tool. All four surfaces below share this one connection.
> ⚠️ **Do this in your normal browser** — not through a remote-desktop session. The OAuth flow is what saves your connection; other login routes can leave Claw reporting "not connected."
### Connect Salesforce from inside Claw (alternative) {#connect-salesforce-from-inside-claw}
If you're already working in Claw, you can authorize Salesforce from the Claw **Services** tab — it's the same one-time OAuth connection, just a different entry point.
> **Before you start:** your Cloud Computer must be running. If it isn't, open the **Settings** tab and start (or create) it first.
- **Step 1 — Open the Services tab:** In Claw, go to the **Services** tab.
- **Step 2 — Find Salesforce under "More Services":** Scroll to the bottom and expand **More Services** — Salesforce won't appear in the main service list, only in this collapsed group.
- **Step 3 — Enable Salesforce:** Click the toggle next to Salesforce. This opens the standard Salesforce OAuth login in your normal browser. Sign in, approve access, and the toggle switches on.
> ⚠️ **Use the toggle, not "Open Remote Desktop"** — the Services tab shows a prompt: *"Don't see what you need? Open Remote Desktop and log in directly."* Do **not** use that path for Salesforce — it bypasses the OAuth token store and will leave Claw reporting "not connected." Use the toggle in Step 3 instead.
The connection is now saved to your Genspark account and shared across every surface.
### How to Verify It Worked {#how-to-verify-it-worked}
You'll know the connection is live when:
- Salesforce shows as **connected** under **Skills → Connectors** in your Genspark account.
- Ask Claw or Super Agent a Salesforce question — for example, *"list my open opportunities"* — and it returns your records.
## Use Salesforce from Claw {#use-salesforce-from-claw data-section-type=feature}
**Best for:** conversational, on-demand Salesforce work.
Once Salesforce is connected, just talk to Claw in plain language. Claw manages the session for you — no manual setup beyond the one-time OAuth connection.
**Plan required:** Claw needs a Genspark Cloud Computer subscription (always-on cloud) or the local Claw desktop app.
### Quick Start {#claw-quick-start}
- **Step 1 — Connect Salesforce:** Authorize Salesforce via OAuth (see section 1).
- **Step 2 — Open Claw:** Open Claw from the Agents list in Genspark, or go to `genspark.ai/claw`.
- **Step 3 — Give instructions in plain language:** For example: *"Show me all open opportunities over $50k closing this month."*
### What You Can Do {#claw-what-you-can-do}
- Pull contacts and leads.
- Update records — change deal stages, log activities, edit fields.
- Query your pipeline.
- Create records (leads, contacts, opportunities).
- Generate reports and summaries — *"Draft a weekly deal summary."*
- Run Salesforce CLI operations including metadata commands (deploy, retrieve, describe) via Genspark's managed CLI sandbox.
## Use Salesforce from Super Agent {#use-salesforce-from-super-agent data-section-type=feature}
**Best for:** multi-step workflows that combine Salesforce with other tasks.
Super Agent works with Salesforce through Genspark's native connector — the same OAuth connection you set up above. Just prompt it in plain language.
### Quick Start {#super-agent-quick-start}
- **Step 1 — Connect Salesforce:** Authorize Salesforce via OAuth (see section 1).
- **Step 2 — Open Super Agent:** Start a Super Agent session.
- **Step 3 — Prompt in plain language:** Super Agent queries and acts on your Salesforce data as it works.
### What You Can Do {#super-agent-what-you-can-do}
- Pull contacts and leads.
- Update records — change deal stages, log activities, edit fields.
- Query your pipeline.
- Create records (leads, contacts, opportunities).
- Generate reports and summaries.
- Combine Salesforce steps with other work — for example, pull at-risk opportunities, draft follow-up emails, and summarize the result in one task.
## Use Salesforce from Workflow Builder {#use-salesforce-from-workflow-builder data-section-type=feature}
**Best for:** repeatable, scheduled automations with specific triggers.
Workflow Builder supports Salesforce operations as natural-language workflow steps — the same way it works with Gmail, Slack, and other tools.
### Actions Available {#workflow-builder-actions-available}
| Action | What it does |
|---|---|
| Create record | Create a Lead, Contact, or Opportunity |
| Update record | Change fields on existing records |
| Query records | Pull data with SOQL |
Use these for scheduled deal-summary reports, automated record creation from other systems, or ongoing pipeline monitoring.
## Use Salesforce from AgentBase {#use-salesforce-from-agentbase data-section-type=feature}
**Best for:** building custom dashboards, views, and reports on your Salesforce data.
Unlike the other surfaces, which act on Salesforce in place, AgentBase brings your Salesforce data into a workspace and lets you build on top of it. It connects as a data source, reads your Salesforce structure and organizes your fields automatically — then you turn that data into views, dashboards, and automations.
### What You Can Do {#agentbase-what-you-can-do}
- Write updates back to Salesforce.
- Build Table, Kanban, Gallery, Calendar, and Form views.
- Create KPI dashboards and charts.
- Run plain-language queries across your imported Salesforce data.
- Set up automations — sync data, generate weekly deal summaries, flag at-risk records, and draft follow-ups.
## Troubleshooting {#troubleshooting data-section-type=troubleshooting}
### Claw says "not connected" after I logged in — what happened? {#claw-not-connected}
You may have logged in through a remote-desktop session instead of the standard OAuth flow. The remote-desktop path bypasses the OAuth token store, so Claw still reports "not connected." Reconnect Salesforce via **Skills → Connectors → CRM & Sales → Salesforce → Connect**, and complete the login in your normal browser.
### What should I check first if a Salesforce request isn't working? {#salesforce-request-not-working}
- Confirm Salesforce still shows as connected under **Skills → Connectors**.
- Confirm the records you're asking for are ones your Salesforce account has permission to see.
- Re-run the request in plain language with specific criteria (object, field, date range).
- If the issue persists, contact Genspark support.
## FAQ {#faq data-section-type=faq}
### Does Salesforce Pardot / Marketing Cloud work? {#salesforce-pardot-marketing-cloud}
Not currently. Standard Salesforce CRM objects — contacts, leads, opportunities, accounts, cases, campaigns, activities, and custom objects — are supported.
### What Salesforce data can Genspark access? {#what-data-can-genspark-access}
All standard CRM objects are available on all surfaces — contacts, leads, opportunities, accounts, cases, campaigns, activities, tasks, and custom objects. Access follows your own Salesforce account's permissions.
For Claw specifically: Claw can also run Salesforce CLI operations including metadata commands (deploy, retrieve, describe) via Genspark's managed CLI sandbox. For Workflow Builder and AgentBase, access is limited to standard and custom record objects — org metadata and admin configuration are not accessible. Super Agent's metadata access scope is unconfirmed.
### Which plan do I need? {#which-plan-do-i-need}
- Claw requires a Genspark Cloud Computer subscription (for always-on cloud mode) or the local desktop app. Claw features also require credits — when credits run out, features like chat and scheduled tasks pause.
- Super Agent, Workflow Builder, and AgentBase work with a standard Genspark account.
# Genspark for Microsoft Office {#genspark-for-microsoft-office data-doc-group=office-suite}
Source: https://www.genspark.ai/helpcenter/microsoft-office-plugin
## What is Genspark for Office? {#overview data-section-type=overview}
Bring Genspark's AI right into the Office apps you already use. Genspark for Microsoft Office is a set of three add-ins — for Word, Excel, and PowerPoint — that open an AI task pane inside the app, so you can draft, analyze, and design without switching tools or copying content back and forth.
> **Genspark for Office vs. GenOffice:** This article is about the Office add-ins that run inside Microsoft Word, Excel, and PowerPoint. That's a different product from GenOffice, our standalone open-source office suite. If you're working inside Microsoft Office, you're in the right place.
Genspark for Microsoft Office brings Genspark's AI into the desktop and web versions of your Office apps as a task pane you open right next to your document.
**What It Does**
Three separate add-ins — one each for Word, Excel, and PowerPoint — let you draft, edit, analyze, and design with AI without leaving the app or copying content elsewhere.
**Who It's For**
Anyone who works in Microsoft Office and wants AI help in place — writing and editing in Word, analyzing and visualizing data in Excel, or building on-brand decks in PowerPoint.
---
## Before You Start {#before-you-start data-section-type=overview}
### Prerequisites {#prerequisites}
Before you begin, make sure you have:
- A Microsoft account with Word, Excel, or PowerPoint — on Windows, Mac, or the web.
- A Genspark account — free to create at [genspark.ai](https://www.genspark.ai).
You'll sign in twice: once with Microsoft (to install the add-in) and once with Genspark (to activate the AI features). That dual sign-in is normal — both are covered below.
---
## Install an Add-in {#install-an-add-in data-section-type=feature}
You can install from Microsoft AppSource or from inside the Office app. Both paths land in the same place.
### Option A — From Microsoft AppSource {#option-a-appsource}
**Step 1 — Open the add-in's AppSource listing**
- Word: [marketplace.microsoft.com](https://marketplace.microsoft.com) — search "Genspark for Word" (AppSource ID: wa200010470)
- Excel: [marketplace.microsoft.com](https://marketplace.microsoft.com) — search "Genspark for Excel" (AppSource ID: wa200010472)
- PowerPoint: [marketplace.microsoft.com](https://marketplace.microsoft.com) — search "Genspark for PowerPoint" (AppSource ID: wa200010473)
**Step 2 — Click Get it now**
Complete the short form (name, job title) and approve Genspark's access permissions.
**Step 3 — Sign in with your Microsoft account** when prompted.
**Step 4 — Open the Office app**
The Genspark task pane appears on the right.
**Step 5 — Sign in to Genspark**
The add-in prompts you to sign in to your Genspark account — click **Sign in**, complete authentication in the browser window that opens, then return to the task pane. The add-in activates once you approve access.
### Option B — From inside Word, Excel, or PowerPoint {#option-b-in-app}
1. Go to **Home > Add-ins** (or **File > Get Add-ins**).
2. Click **More Add-ins > Store**.
3. Search "Genspark" and click **Add**.
4. Open the task pane. When prompted to sign in to Genspark, click **Sign in** and complete authentication in the browser window that opens — then return to the task pane.
To manage or remove an add-in later: go to **Home > Add-ins > My Add-ins**, right-click (or select the add-in and click ⋯), and choose **Remove**.
### How Sign-In Works {#how-sign-in-works}
Genspark uses a device-code sign-in: the task pane starts the process, your system browser opens for you to log in and approve, and the add-in polls until you've approved — then it activates. You authenticate in the browser, not inside the task pane (this avoids third-party-cookie limits in Office on the web).
---
## What You Can Do in Word {#word data-section-type=feature}
Open the Genspark AI task pane from the Home tab, then ask in plain language.
- **Draft and edit** — write new sections, expand a rough outline, summarize long passages, or rewrite for tone. Try: *"Rewrite this paragraph to be more concise and confident."*
- **Analyze your document** — get word count, structural analysis, and writing-quality suggestions.
- **Format with words** — apply styles, headings, tables, and lists by describing what you want.
- **Insert math equations** — add native, editable Office Math objects.
- **Generate images** — create illustrations and diagrams directly in the document.
- **Research the web** — run searches without leaving Word.
- **Work in multiple languages** — English, Chinese, and more.
---
## What You Can Do in Excel {#excel data-section-type=feature}
Genspark for Excel is your AI data expert — select a range or just ask from the task pane.
- **Generate formulas** — describe what you need in plain English, from a VLOOKUP to complex arrays. Genspark writes it, explains it, and inserts it into the cell. Try: *"Create a VLOOKUP that pulls the price from the Products sheet by SKU."*
- **Analyze and visualize** — select a data range and get native Excel charts (bar, line, pivot, combo).
- **Get instant insights** — ask a question in plain language and get trends, outliers, and summaries with no formula required.
- **Explain any formula** — paste a formula and get a step-by-step breakdown to debug or improve it.
- **Research into your sheet** — Genspark searches the web, structures the results, and fills in your spreadsheet.
---
## What You Can Do in PowerPoint {#powerpoint data-section-type=feature}
Genspark for PowerPoint helps you build on-brand, presentation-ready decks without leaving the app.
- **Stay on brand** — works with your corporate master slides so every slide matches your template.
- **Structure and lay out** — builds the outline and narrative, then picks the best layouts.
- **Point and edit** — select any element and tell the AI what to change; built-in polishing workflows handle the rest.
- **Draft speaker notes** — notes are written as the deck is designed.
- **Add studio-quality voiceovers** — generate and embed a voiceover in one click.
- **Insert editable charts** — real, native PowerPoint charts built from your data.
- **Create business visuals** — web image search plus AI-generated custom graphics.
---
## Genspark for Office vs. GenOffice {#office-vs-genoffice data-section-type=feature}
Two different products share the "office" name:
- **Genspark for Microsoft Office** (this article) — add-ins that run inside Microsoft Word, Excel, and PowerPoint. A shipped, commercial product.
- **GenOffice** — a standalone, open-source office suite (Docs, Sheets, Slides, PDF) with AI built in.
If you use Microsoft Office, you want the add-ins. If you're looking for a separate, free-standing suite, that's GenOffice.
---
## Troubleshooting {#troubleshooting data-section-type=troubleshooting}
### The Genspark task pane isn't showing up. {#task-pane-not-showing}
Make sure the add-in finished installing (**Home > Add-ins > My Add-ins**) and that you have an active internet connection. Close and reopen the app if the pane doesn't appear.
### I'm stuck signing in to Genspark. {#stuck-signing-in}
There are two sign-ins: your Microsoft account (to install) and your Genspark account (to activate features). When you're prompted to sign in to Genspark, click **Sign in** — authentication opens in your browser, not inside the task pane. Log in there, approve access, and return to the task pane. If nothing seems to happen, check that the browser window or tab didn't open behind your Office window, then try again.
### My organization disabled the add-in. {#organization-disabled-add-in}
Add-ins can be blocked by your company's IT policy or Microsoft Trust Center settings. If you're on a managed device, ask your admin to approve the Genspark add-in from AppSource.
## FAQ {#faq data-section-type=faq}
### How do I remove an add-in? {#how-do-i-remove-an-add-in}
Go to **Home > Add-ins > My Add-ins**, right-click (or select the add-in and click ⋯), and choose **Remove**.
# AI Drive {#ai-drive data-doc-group=knowledge}
Source: https://www.genspark.ai/helpcenter/ai-drive
## What is AI Drive? {#what-is-ai-drive data-section-type=overview}
**AI Drive** is Genspark's built-in cloud drive — a single place to store, organize, and share every file you produce and use across Genspark. The biggest difference from an ordinary cloud drive: all of Genspark's AI products are wired into your drive, so files don't have to be exported and re-imported between products.
- **All your Genspark files in one place** — decks built in Slides, documents written in Docs, material downloaded by the AI, and files produced by AI coding all land in the drive automatically, ready to find again anytime.
- **Files wired directly into the AI** — hover any file and click **Super Agent** to hand it to the AI. Ask it to research the document, turn it into a deck, or generate a video. The output goes right back into AI Drive.
- **Easy to organize** — folders, move, rename, search, and list/grid views are all there; images, PDFs, PPTX, video, and zip archives can be browsed inside without extracting, all without downloading.
- **Built for sharing and collaboration** — share any file or folder by link or email invite, with control over who can view and edit.
- **Recover deleted files** — deleting moves to trash first, where files can be restored within 30 days.
**How it differs from other Genspark products:** AI Chat, Super Agent, and AI Slides do work — they produce results. AI Drive is where you keep things. It's the unified file hub where each product's output can live, and from which you can hand files back to the AI to process further. Unlike external cloud drives like Google Drive or OneDrive, AI Drive is built into Genspark — your files and the AI tools share the same space.
## Quick Start {#quick-start data-section-type=overview}
- **Go to AI Drive** — open **Drive** from the left sidebar (/aidrive/files). The first time you open AI Drive, a legal disclaimer appears — check "I have read and agree" and click **I AGREE** before uploading or downloading. Once you have files, a **Recent Files** section appears at the top of the root.
- **Put files in** — click **"+ New"** at the top right and choose Upload (files) or Upload Folder, or drag files onto the page. Or use **Download For Me** to pull material straight into the drive from a URL or a one-line description. Output from Genspark products lands automatically.
- **Organize, preview, and use** — organize with folders, move, and rename; double-click a file to preview it online; select a file and use the top action bar to share, download, or hand it to the AI.
**Tip** — File-management actions (upload, download, organize, compress, share, delete, and restore) don't consume credits. You're only billed by an AI product's own rules when you hand a file to that AI to process.
## Upload and Organize Files {#upload-and-organize-files data-section-type=feature}
### Upload files {#upload-files}
Click **"+ New"** at the top right and pick from the dropdown:
- **Upload** — upload one or more files
- **Upload Folder** — upload a whole folder (keeps the directory structure)
- **New Folder** — create an empty folder
You can also drag files onto the page. Progress and per-item results appear in the **Status / Operations panel at the bottom right** (e.g. "Successful" or "Moved 1/1 Completed").
**Tip** — Output from Genspark products is saved automatically — if you can't find a generated file, check /aidrive/files first.
### Move, rename, organize {#move-rename-organize}
Select a file or folder and a row of actions appears at the top: **Share / Rename / Download / Move / Delete**.
### Views, sorting, and Recent Files {#views-sorting-and-recent-files}
The file area switches between **list** and **grid** views, and sorts by Name / Size / Modified. **Recent Files** at the top of the root shows recently accessed files in a horizontal row — click one to open it quickly.
### Search {#search}
Use the search box at the top ("Search files and folders") to find files; results return as you type, each labeled with where the file lives.
### Online preview {#online-preview}
Double-click a file to open a full-screen preview. Common types that preview online: images, PDFs, PPTX, and video. Zip archives can be browsed inside without extracting. Image previews can be **cropped/rotated** and saved back to Drive. Types that can't preview online prompt you to download.
## Download Into the Drive {#download-into-the-drive data-section-type=feature}
No need to download to your computer and re-upload — paste a link or describe what you want, and the AI downloads it straight into the drive.
### Download For Me {#download-for-me}
Click **"Download For Me"** in the drive to open the downloader:
- **Tell it what to download** — paste an HTTP/HTTPS link or describe what you want in natural language → click **Confirm**.
- **Pick files and save** — in "Select Files to Download," check the files to keep, choose a destination, and click **"Save to AI Drive"**.
**Tip** — Common web resources — video, audio, images, documents — can all be downloaded. Downloading a direct link doesn't consume credits; any exceptions follow what the UI shows.
### Compress and decompress {#compress-and-decompress}
Select files or folders → **Compress** (creates a .zip) or **Decompress** (extracts a .zip into a folder). Runs as a background task with progress shown.
### Download to your computer {#download-to-your-computer}
Select a file → click **"Download"**. For multiple files, confirm **"Compress and Download"** to bundle them into a single archive.
## Delete and Trash {#delete-and-trash data-section-type=feature}
Deleting isn't gone-forever — files go to trash first, giving you a chance to change your mind.
### Delete to trash {#delete-to-trash}
Select a file → click **"Delete"** → confirm "Move to trash." Files move to trash rather than being permanently deleted. Auto-cleared after **30 days**. Deleting a folder removes every file inside.
### Restore or empty the trash {#restore-or-empty-the-trash}
Open trash at /aidrive/trash. Select items to **Restore** or **Delete** just those. With nothing selected, the buttons become **Restore All** and **Delete All** for the whole trash. If a file's original folder is gone, those items are listed by name with a count of how many need a new location chosen. If some items fail to restore for another reason, you'll see "Restored X of Y items (N failed)."
## Share Files and Folders {#share-files-and-folders data-section-type=feature}
Share any file or folder and control exactly who can open it and what they can do. Sharing works two ways you combine: **invite specific people** (by email, with a role) and **set a general-access level** (how broadly a link works). Start by selecting a file or folder and clicking **Share**.
**Drive has two main tabs:**
- **My Drive** — all the files and folders you own
- **Shared with me** — files and folders others have shared with you. The file stays in the **owner's** storage (doesn't use your space). If you have edit access and delete something from a shared folder, it goes to the **owner's Trash** — the owner can restore it.
### Inviting people {#inviting-people}
In the Share panel, use the **"Add by email"** field to add people, then choose their role (default: **Can edit**):
| Role | What they can do |
| --- | --- |
| **Can view** | Open and read (and download, when allowed) |
| **Can edit** | Everything a viewer can do, plus make changes |
Click **Share** to send the invite. Invited people access the file via their Genspark account — the invite is tied to their account, so they'll sign in the first time they open it.
**Manager** is a separate governance role for managed or team folders — not something you assign from this panel. It controls the access list, settings, and trash for the folder.
**Note on org drives** — Administrators of an organization are **implicit Managers of that organization's own drive** — admin edit and read access even without appearing in a file's share list. Resolved live, never stored. **Does not extend to members' personal drives.**
### General access: how broadly the link works {#general-access-how-broadly-the-link-works}
- **Restricted** — only explicitly invited people can open it. This is the default.
- **Anyone in your organization** (or "Anyone in [org name]") — anyone in your Genspark org can open it.
- **Anyone with the link** — anyone who has the link can open it, **without signing in**. A signed-out visitor gets a read-only view. This access level is **view-only** — the role is locked to view; only Restricted and Organization access can be Editor.
**Tip** — **Allow downloads:** the "Let viewers download the file" (or "Let viewers download files in this folder") toggle is in the Share panel and controls whether the people you've shared with can download — or only view. It applies to all share recipients, including email invitees.
**Note on org policy** — Your org's policy controls whether public sharing is allowed. If "Anyone with the link" isn't offered, an org policy is blocking it. If an org tightens its policy later, **links already shared publicly stop working retroactively.** When org policy blocks all public sharing, you'll see "Your organization policy restricts public AI Drive sharing" and only email invites work.
### Changing or revoking access {#changing-or-revoking-access}
Open the Share panel on a shared item:
- To remove a person's access: click **Remove** on their row
- To stop link sharing: set General access back to **Restricted**
## Let the AI Work on Your Drive Files {#let-the-ai-work-on-your-drive-files data-section-type=feature}
Files in the drive aren't just sitting there — hand one to the AI to keep working on with one click, and the output goes back into the drive.
### Start from a file with Super Agent {#start-from-a-file-with-super-agent}
Hover over any file row → click **"Super Agent"** → Super Agent opens in a new tab carrying that file. Type your request — research, organizing, turning it into a deck, generating a video — and the output is saved back to Drive.
Folders and 0-byte empty files have the Super Agent button grayed out. To use a specific product directly (Slides, Docs, etc.), go into that product and pick a file there.
### Agents read and write your drive directly {#agents-read-and-write-your-drive-directly}
Genspark's agents (Super Agent, AI coding, etc.) can read and write your drive directly while running a task — files produced by AI coding, downloaded material, and generated reports all land in Drive and can be reused across tasks.
**About cost:** keeping files in Drive is free. The moment you hand a file to an AI like Super Agent, Slides, or Deep Research to process, it's billed by that product's rules — not by Drive. The drive's core file management stays free.
## Troubleshooting {#troubleshooting data-section-type=troubleshooting}
### Why can't I choose "anyone with the link"? {#anyone-with-link-unavailable}
Your org's policy may block public sharing. You'll see "Your organization policy restricts public AI Drive sharing" — only email invites work in that case. Contact your org administrator.
## FAQ {#faq data-section-type=faq}
### Is AI Drive the same as Google Drive or OneDrive? {#vs-google-drive-onedrive}
No. AI Drive is Genspark's own cloud drive, deeply wired into Genspark's AI products. You can import files from Google Drive / OneDrive to use in Genspark, but those are separate external cloud drives.
### Can free users use it? {#free-users}
Yes. Drive is open to all users. File-management actions are free; storage space varies by plan (see the usage bar at the top).
### How big is the drive? Does it cost money? {#storage-size-cost}
File management itself is free — upload, download, organize, compress, share, delete, and restore don't consume credits. Storage space is a membership benefit; higher plans get more, and the exact capacity is shown in the usage bar at the top. When space is full, uploads are blocked with a prompt to upgrade.
### Do files generated by Slides / Docs save automatically? {#auto-save-generated-files}
Yes. Output from Genspark products lands in your drive automatically at /aidrive/files. If you can't find a generated file, check here first.
### Can I recover something I deleted by mistake? {#recover-deleted-files}
Yes. Go to /aidrive/trash — select individual items and click
**Restore**, or click **Restore All** with nothing selected to restore everything. If a file's original folder no longer exists, those items are listed by name so you can choose a new location. If some items fail for another reason, you'll see "Restored X of Y items (N failed)." After emptying the trash ("Delete All") or after 30-day auto-clear, recovery isn't possible.
### How do I control who can see or change a shared file? {#control-share-access}
Use the Share panel: add people by email (Can view or Can edit), and set the general-access level (Restricted / Anyone in your org / Anyone with the link). "Anyone with the link" is always view-only — only email-invited people can be given edit access.
### Can I stop sharing after I've sent a link out? {#stop-sharing}
Yes. Open the Share panel: click **Remove** on a person's row to remove their access, or set General access back to **Restricted** to stop link sharing. After you set it back to Restricted, the link stops working immediately.
### Does "anyone with the link" require a Genspark account? {#anyone-with-link-account-required}
No — a signed-out visitor gets a read-only view without signing in. Invite-by-email access does require an account.
### Can an administrator see my org's drive files? {#admin-visibility}
Admins are implicit Managers of the organization's own drive — edit and read access even without appearing in a file's share list. Does not extend to members' personal drives.
### Where do files people share with me appear? {#shared-with-me-location}
In the **Shared with me** tab. They stay in the owner's storage (don't count against your space). If you have edit access and delete something, it goes to the **owner's Trash** — the owner can restore it.
### Which files can use the Super Agent button? {#super-agent-button-eligibility}
All non-folder files can (hover a file row and the Super Agent button appears). Folders and 0-byte empty files show the button grayed out.
### Does letting the AI work on drive files cost money? {#ai-processing-cost}
Keeping files in Drive is free. Processing a file with an AI (Super Agent, Slides, etc.) is billed by that product's rules.
### What can Download For Me download? {#download-for-me-scope}
Common web resources — video, audio, images, documents — by pasting a link or describing in natural language. Downloading a direct link into the drive doesn't consume credits.
# Gen-1 {#gen-1 data-doc-group=office-suite}
Source: https://www.genspark.ai/helpcenter/gen-one-model
## What is Gen-1? {#what-is-gen-1 data-section-type=overview}
**Gen-1** turns a single request into a boardroom ready presentation and is Genspark's first in-house model trained to frontier quality as a new model for **AI Slides standard mode**.
Gen-1 is built from **MiniMax M3**, an open-weight model, and was post-trained with reinforcement learning inside the same production system that runs AI Slides. Gen-1 ranks first in 8 of 9 internal and public benchmark comparisons, putting it in the same tier as Claude Opus 5.
## Where you'll find Gen-1 {#where-youll-find-gen-1 data-section-type=feature}
Gen-1 is the **standard model inside AI Slides** to build your deck. No setup or selection needed. Standard and Ultra are chosen at the bottom of the AI Slides chat section.
- **Standard mode** — runs Gen-1 by default (included in every plan)
- **Ultra mode** — keeps offering frontier models (top-tier models from other labs) such as **Claude Opus 5**
Gen-1 is additive: it joins the models Genspark already orchestrates. Standard mode leads with Gen-1, and Ultra mode keeps the frontier models available whenever you want them.
## What you get with Gen-1 {#what-you-get-with-gen-1 data-section-type=feature}
Gen-1 specializes in turning a request into a deck you can use as-is. Giving you the output quality of ‘**Ultra**’ in our ‘**Standard**’ mode.
- **Polished visual design** — trained to match the quality of frontier-tier decks
- **Clean, reliable layouts** — structured slides built end to end, ready for your audience
- **End-to-end slide building** — planning, design, and iteration handled from a single prompt
## Privacy & your data {#privacy-your-data data-section-type=feature}
Running on USA infrastructure, your data never leaves your environment and is never shared with the base model’s developer. MiniMax is not a subprocessor and has no access to your data. This is the same boundary applied to every model used.
## Cost {#cost data-section-type=feature}
- **Gen-1 is now the model used for** Standard mode in AI Slides, is included in every plan at no change in price but with better results.
- For teams wanting **API access** to this model please reach out to modelpartner@genspark.ai. For existing enterprise customers who have workspace model restrictions, this remains unchanged. Standard mode continues to use approved models as before, workspaces can enable or disable Gen-1 slides on request.
## FAQ {#faq data-section-type=faq}
### Is Gen-1 open-weight? {#open-weight}
No. It's built on the open-weight model **MiniMax M3**, but the post-trained Gen-1 model itself is not open-weight.
### Does this mean Genspark is done orchestrating other models? {#still-orchestrating-other-models}
No. Genspark continues to orchestrate the best models for every task, frontier and open alike. Gen-1 is the first model trained where our scale and product experience give us an advantage — slides lead the way.
### Was Gen-1 trained on my decks? {#trained-on-my-decks}
No. User content was not used as training data; only internally constructed tasks were. Aggregated, de-identified feedback is used to calibrate the grader, not to train the model.
### Does Gen-1 cost extra? {#cost-extra}
No — it's the default in Standard mode of AI Slides, which is included in every plan at no change in cost.
### Can I use other models for slides? {#other-models-for-slides}
Yes. Ultra mode in AI Slides continues to offer frontier models like **Claude Opus 5**.
### What if my deck doesn't look right? {#deck-doesnt-look-right}
Just describe the change and it's regenerated, or adjust any slide directly — decks are fully editable. If you'd rather start from a different model, you can switch to Ultra mode at any time.
# Genspark Super App {#genspark-super-app data-doc-group=build-suite}
Source: https://www.genspark.ai/helpcenter/super-app
## What is the Genspark Super App? {#what-is-the-genspark-super-app data-section-type=overview}
Genspark Super App brings Super Agent, GenMail, GenTeam, and GenCode together in one desktop app for macOS, Windows, and Linux.
Four products. One window. Each product lives as its own tab:
| Tab | What it is |
| --- | --- |
| Super Agent | Chat and run tasks in a cloud sandbox |
| GenMail | A full email and calendar client |
| GenTeam | The team workspace |
| GenCode | Genspark's coding agent |
Everything shares one sign-in, so you switch between your AI workspace, your inbox, and your team in one space.
### Why use the Super App? {#why-use-the-super-app}
Everything you can do on genspark.ai is here too, and stays in sync. The Super App adds what only a real desktop app can:
- **One notification center** — every tab's alerts in one place: per-tab switches, sound control, a bell with history, and a single accurate dock badge
- **One sign-in** — sign in once, every tab shares the session; credentials are encrypted at rest and never exposed to page content; enterprise SAML SSO is supported
- **Stable and beta channels** — safe auto-updates on the stable channel, or opt into beta and get every improvement as it ships
## Quick Start {#quick-start data-section-type=overview}
1. Go to [superapp.genspark.ai](https://superapp.genspark.ai)
2. Choose your platform: macOS (Apple Silicon), Windows, or Linux
3. Install and sign in with your Genspark account
It's free to download. Mac and Windows builds are signed; Linux ships as an AppImage. One sign-in works for every tab, you don't create a separate account per product.
The Mac build supports Apple Silicon machines.
## Feature & How-To Guides {#feature-and-how-to-guides data-section-type=feature}
### Super Agent {#super-agent}
The Super Agent in its own tab:
- Full cloud task list, always in sync with genspark.ai
- Live tool calls and artifacts inside the conversation
- The agent can work on this computer's files
### GenMail {#genmail}
A complete mail client:
- Fast reading, compose, and full-text search
- Calendar with scheduling
- In-pane attachment preview and mail translation
- Email Brain and a skills manager for mail automation
### GenTeam {#genteam}
The team workspace, native in its own tab:
- Channels, direct messages, and a full rich-text composer
- Built-in viewers for PDF, Word, code, CSV, and archives
- Computers pane for the team's cloud machines
### GenCode {#gencode}
Gensparks coding agent, in its own space:
- Works on the code in a directory you choose
- Runs actions as tool calls, returning text and file changes
- Handles larger tasks through read-only and read-write sub-agents
## Desktop vs web {#desktop-vs-web data-section-type=feature}
| What | Web (genspark.ai) | Super App |
| --- | --- | --- |
| Super Agent | ✅ | ✅ — cloud task list and Sandbox runs stay in sync |
| GenMail | ✅ | ✅ (native tab) |
| GenTeam | ✅ | ✅ (native tab) |
| One notification center + dock badge | — | ✅ |
| Multiple product tabs in one window | — | ✅ |
| Native app qualities (notifications, auto-updates) | — | ✅ |
Nothing is exclusive to the web — your accounts, tasks, and projects carry over through the same sign-in, so you can switch between the app and the browser as you like.
## FAQ {#faq data-section-type=faq}
### Is the Genspark Super App free? {#is-the-genspark-super-app-free}
Yes. It's free to download. Sign in with your Genspark account — usage beyond that follows your normal Genspark plan and credits.
### Do the Super App and genspark.ai stay in sync? {#do-the-super-app-and-genspark-ai-stay-in-sync}
Yes. Your conversations and your cloud task list stay in sync, so you can switch between the app and the web without losing context.
### Do I need a separate account for each tab? {#do-i-need-a-separate-account-for-each-tab}
No. One Genspark sign-in covers all four tabs. Credentials are encrypted at rest and shared across tabs without being exposed to page content.
### Which Macs are supported? {#which-macs-are-supported}
The Mac build is for Apple Silicon machines.
### Is there a Windows or Linux build? {#is-there-a-windows-or-linux-build}
Yes — Windows and Linux are both supported; Linux ships as an AppImage.
### Does the Super App support enterprise sign-in? {#does-the-super-app-support-enterprise-sign-in}
Yes. Enterprise SAML SSO is supported, alongside regular Genspark account sign-in.
### Is GenMail in the Super App the same as the GenMail app? {#is-genmail-in-the-super-app-the-same-as-the-genmail-app}
Yes — GenMail in the Super App is the same GenMail product, hosted as a tab in the app. The tab supports the core GenMail experience (compose, Email Brain, snooze, translate, calendar booking, free/busy) rather than being a webview. Two caveats: the Super App and the standalone GenMail app support different platforms (see [Quick Start](#quick-start)), and whether the tab includes every standalone feature — for example calendar sharing and agentic search — is still being confirmed with product.
# GenCode {#gencode data-doc-group=build-suite}
Source: https://www.genspark.ai/helpcenter/gencode
## What is GenCode? {#what-is-gencode data-section-type=overview}
GenCode is Genspark's coding agent. It works on code in a directory you choose. Pick between a Chat UI and a Terminal UI — the same product with the same capabilities, whichever style you prefer — and you can run it in two places:
- **The GenCode workspace in the Genspark Super App** — the desktop surface. You work with the agent in a conversation, and the prompt composer shows the model in use, your context usage, and credits.
- **The standalone GenCode CLI** — a terminal program you install with npm and run from the directory you want to work in.
Both run the same engine, the same Genspark-hosted model catalog, and the same billing. GenCode is built on Genspark's fork of OpenCode, and it keeps its own configuration directories, so it does not touch an existing OpenCode install.
## Quick Start {#quick-start data-section-type=overview}
### In the Genspark Super App {#in-the-genspark-super-app}
1. Update the Genspark Super App to the latest version.
2. Open the app and look for the GenCode workspace. If you don't see it, click **+** in the tab bar and add it.
No Super App yet? Download it [here](https://superapp.genspark.ai/). It is a free download for macOS, Windows, and Linux. On Mac, the Super App requires Apple Silicon.
### With the CLI {#with-the-cli}
1. **Install it:** `npm i -g @genspark/gencode`
2. **Sign in:** `gencode login` — this opens a device-flow sign-in. Once signed in, the CLI shows your balance, plan, and account email next to the model catalog.
3. **Run it:** `gencode`, from the directory you want to work in.
The CLI keeps the OpenCode command surface:
- `gencode run --format json` for scripted and non-interactive use
- `gencode models` to list the models available to you
- `--dir`, `--session`, and `--model` flags
- environment variables prefixed `OPENCODE_`
It keeps its own configuration directories — `~/.config/gencode`, `~/.local/share/gencode`, `~/.local/state/gencode`, and `~/.cache/gencode` — so an existing OpenCode install is untouched.
The CLI binaries cover Intel Macs as well as Apple Silicon Macs, plus Windows and Linux.
## Feature & How-To Guides {#feature-and-how-to-guides data-section-type=feature}
GenCode ships with a Chat UI and a Terminal UI. It is the same product with the same capabilities either way — use the one that fits how you work.
- **The workspace.** In the Genspark Super App, you work with GenCode in a conversation. The composer shows the model in use, your context usage, and credits.
- **The CLI.** The same agent in your terminal, as a full-screen interface with panels for session context and code diagnostics.
- **A coding agent either way.** GenCode works on the code in a directory you choose. Actions run through tool calls, and the results are text and file changes.
- **Sub-agents.** GenCode handles larger tasks through sub-agents — one that reads and analyzes, and one that can make changes. Each has a configured step limit, so a run stays bounded.
- **Ratings.** In the Super App, you can rate any GenCode turn with a thumbs up or thumbs down. Ratings stay on your machine.
- **Your feedback stays local.** A rating describes the turn it came from — model, tools used, files and lines changed, tokens, cost, and time. Any free-text comment you write is never transmitted.
- **Scriptable and CI-friendly.** `gencode run --format json` returns machine-readable output you can pipe into scripts and pipelines.
## Models & cost {#models-and-cost data-section-type=feature}
**Models.** GenCode uses a Genspark-hosted model catalog spanning nine model families: Claude (Anthropic), GPT (OpenAI), Gemini (Google), DeepSeek, GLM (Z.ai), MiniMax, Kimi (Moonshot AI), Grok (xAI), and Nemotron (NVIDIA).
**Cost.** GenCode runs on Genspark credits. Usage is billed to the account GenCode authenticates with. In the CLI, if `GSK_API_KEY` (or `GSK_BASE_URL`) is set in your environment, the key's account is billed instead. In your credit history, GenCode usage appears as **"GenCode"**.
GenCode is billed straight to credits, rather than through the Free Quota that covers Standard-mode agent tasks — see [Membership Plans](https://www.genspark.ai/helpcenter/membership-plans).
There is no separate GenCode subscription or wallet — it uses your existing Genspark account and credits. For how credits work, see the [Credits Guide](https://www.genspark.ai/helpcenter/credits-guide).
## FAQ {#faq data-section-type=faq}
### Is GenCode the same as Genspark Code? {#is-gencode-the-same-as-genspark-code}
No. Genspark Code is the browser agent at genspark.ai/code that builds complete applications from a prompt — see the [Genspark Code help article](https://www.genspark.ai/helpcenter/ai-developer). GenCode is the coding agent described here.
### Is GenCode the same as Genspark Claw? {#is-gencode-the-same-as-genspark-claw}
No. Claw is Genspark's personal-assistant agent that operates your apps and files, not a coding agent.
### What about GenTerminal? {#what-about-genterminal}
GenTerminal is Genspark's standalone terminal app — a different product.
### Do I need the Super App to use GenCode? {#do-i-need-the-super-app-to-use-gencode}
No. You can use the GenCode workspace in the Genspark Super App, or install the standalone GenCode CLI. Both are the same agent.
### What is the difference between the workspace and the CLI? {#what-is-the-difference-between-the-workspace-and-the-cli}
Just where you run it. The workspace runs inside the Genspark Super App; the CLI runs in your terminal. Same agent, same capabilities.
### Do I need a paid plan? {#do-i-need-a-paid-plan}
No. GenCode uses your existing Genspark account and credits. There is no separate GenCode subscription or wallet.
### Does GenCode cost credits? {#does-gencode-cost-credits}
Yes. GenCode is billed to the account it authenticates with — by default the account you signed in with. In the CLI, if `GSK_API_KEY` is set in your environment, the key's account is billed instead. Usage appears as "GenCode" in your credit history, and GenCode is billed straight to credits rather than through Free Quota.
### Which models does GenCode use? {#which-models-does-gencode-use}
A Genspark-hosted catalog spanning nine model families: Claude (Anthropic), GPT (OpenAI), Gemini (Google), DeepSeek, GLM (Z.ai), MiniMax, Kimi (Moonshot AI), Grok (xAI), and Nemotron (NVIDIA). Which specific models your account can use depends on the live catalog and your organization's AI-vendor policy.
### Will the CLI interfere with my existing OpenCode install? {#will-the-cli-interfere-with-my-existing-opencode-install}
No. GenCode is built on Genspark's fork of OpenCode, but it installs its own binary and keeps its own configuration directories (`~/.config/gencode`, `~/.local/share/gencode`, `~/.local/state/gencode`, `~/.cache/gencode`). It never touches an existing OpenCode install.
### Does GenCode work on an Intel Mac? {#does-gencode-work-on-an-intel-mac}
The CLI does — its published npm packages include Intel macOS (darwin-x64). The Super App requires Apple Silicon on Mac. Windows and Linux are supported for both.
### Where do I download the Genspark Super App? {#where-do-i-download-the-genspark-super-app}
To download Super App follow this [link](https://superapp.genspark.ai/) — free download for macOS, Windows, and Linux.
# AI Chat {#ai-chat data-doc-group=content-creation}
Source: https://www.genspark.ai/helpcenter/ai-chat
## What is AI Chat? {#what-is-ai-chat data-section-type=overview}
AI Chat is Genspark's conversational agent. It answers questions, can search the web, reads uploaded documents, and writes or edits text. It produces text only.
Use AI Chat to shape and refine what you want: talk an idea through, tighten the ask, get the thinking done, then hand the finished prompt to an agent to execute. Nail the prompt in AI Chat first, then let Super Agent run with it.
## Quick Start {#quick-start data-section-type=overview}
1. Open AI Chat — from the Genspark homepage, or go straight to the AI Chat agent. It's also available in mobile browsers and the mobile apps.
2. Type your question.
3. Pick a model — or use Mixture-of-Agents (below).
4. Follow up — keep chatting in the same conversation, and switch models mid-conversation whenever you like.
## Feature & How-To Guides {#feature-and-how-to-guides data-section-type=feature}
### Model picker {#model-picker}
The model picker is a single control — there are no Standard or Ultra tiers in AI Chat. It lists 40+ models plus a Mixture-of-Agents option.
### Mixture-of-Agents {#mixture-of-agents}
Auto-mixes best AI models for your task. An ensemble of models whose outputs are synthesized into a single response.
### Web search with sources {#web-search-with-sources}
Toggle Search Web and AI Chat searches the web in real time. Results include links to the sources so you can verify what it says. Search is disabled on a few specific models (see the list in AI Chat for the most up to date models).
### Files {#files}
Upload PDF, Word, Excel, and image files and AI Chat reads them. Some models also accept image input directly.
### Personalization {#personalization}
Personalization is on by default, and you can turn it off from the app.
### Product documentation {#product-documentation}
Genspark's product documentation is always attached, so AI Chat can answer questions about Genspark products.
## Plans & credits {#plans-and-credits data-section-type=feature}
- **Free** — a daily credit grant (100 credits a day up until your lifetime free credit allowance is used up). Chat is limited by your credit balance, not by which model you pick. Claude Haiku 4.5 is the default model.
- **Plus and Pro** — Core models are included at no credit cost within the fair use limit before consuming your monthly credits. When you reach the fair use limit, you continue to use your monthly credit balance.
- **Plus and Pro, subscribed before September 18, 2026** — Your current unlimited zero credit benefits continue unchanged through December 31, 2026 (continuously active subscription required): every model included at no credit cost.
- **Team** — Team seats are included in this benefit set; seats active before September 18, 2026 and kept continuously active retain the unlimited chat and image benefit through December 31, 2026.
### Fair-use cap {#fair-use-cap}
To keep Genspark fast, fair, and reliable, paid plans have a fair-use cap on AI Chat, separate from your credit balance: a session limit over a fixed 5-hour window. It's per-user and anchored to your first message — new chats don't reset it, and buying credits doesn't lift it.
When the cap is reached, AI Chat shows you exactly when it will be available again — "AI Chat [5-hour limit] reached. Resets {time}" while all other agents keep working as usual.
If you subscribed before September 18, 2026 and keep your subscription continuously active, your current benefits continue unchanged through December 31, 2026.
## FAQ {#faq data-section-type=faq}
### What is AI Chat? {#faq-what-is-ai-chat}
A general-purpose conversational assistant — multi-model text chat. It answers questions, researches with web search, reads uploaded documents, and writes or edits text.
### How is AI Chat different from Super Agent? {#how-is-ai-chat-different-from-super-agent}
Use AI Chat to shape and refine what you want — talk an idea through and tighten the ask. Then hand the finished prompt to Super Agent to execute.
### How do I choose a model? {#how-do-i-choose-a-model}
Use the model picker. It lists 40+ models plus Mixture-of-Agents, and it's the source of truth for what's available — the lineup changes often.
### What is Mixture-of-Agents? {#what-is-mixture-of-agents}
It auto-mixes best AI models for your task, synthesizing their outputs into one response.
### Can AI Chat generate images? {#can-ai-chat-generate-images}
No — AI Chat produces text only. For images, use the AI Image Agent.
### Can I use voice input in AI Chat? {#can-i-use-voice-input-in-ai-chat}
No — voice input isn't part of AI Chat. Dictation is a separate product.
### Can AI Chat read my files? {#can-ai-chat-read-my-files}
Yes — upload PDF, Word, Excel, or image files and it reads them. Some models also accept image input directly.
### Does AI Chat remember my past conversations? {#does-ai-chat-remember-my-past-conversations}
No — AI Chat doesn't remember you across conversations. It has no memory system and doesn't read your Hub files. Your chats are saved as projects, so ending a conversation doesn't delete it.
### Why did I hit a limit even though I have credits? {#why-did-i-hit-a-limit-even-though-i-have-credits}
That's the fair-use cap — a token-spend limit over a fixed 5-hour window, separate from your credit balance. It resets when the window ends; buying credits doesn't lift it. See [Membership Plans](https://www.genspark.ai/helpcenter/membership-plans) for details.
### How much does AI Chat cost? {#how-much-does-ai-chat-cost}
On the free plan, chat is limited by your daily credit allowance. On Plus or Pro plans, core models are included at no credit cost within the fair use limit before consuming credits and flagship models use credits. When you reach the limit you continue with credits purchased rather than waiting.
### How do I turn off personalization? {#how-do-i-turn-off-personalization}
Personalization is on by default; you can turn it off from the app.
### What about my data? {#what-about-my-data}
For data retention, see the AI Data Retention setting under Settings → Data Controls.
# Account Management
Source: https://www.genspark.ai/helpcenter/account-management
## FAQ {#faq data-section-type=faq}
### What should I do if I can't log in? {#cant-login}
**If you cannot log in, try these steps:**
1. **Verify your registration email** — Use the email from your original account creation, not your billing email (i.e., the email you entered on Stripe to receive invoices)
2. **Check for typos** — Carefully review your email spelling
3. **Check spam folder** — Verification or password reset emails may be filtered
4. **Contact support** — If issues persist, [submit a ticket](#support){.helpcenter-support-link}
### What are the important things to remember about my login email? {#login-email}
**Key points to remember:**
- Your login email and billing email may be different (i.e., your billing email is the email you entered on Stripe for invoices)
- Do not use your billing email to log in
- Always use the email address you registered with
### How do I reset my password? {#reset-password}
**Step-by-step process:**
1. Click "Forgot Password" on the login page
2. Enter your registration email
3. Follow the reset link sent to your email
4. Create a new password
**Security Note:**
- For security reasons, Genspark support cannot access or directly change your password
- All password resets must be done through the self-service "Forgot Password" system
- This is the only way to change your password
### What should I know about logging in with Google? {#google-login}
**Important for Google users:**
- If you created your account using Google Sign-In, you cannot set a separate password
- Each time you log in, continue using "Sign in with Google" to access your account
- This is a platform limitation for security and authentication purposes
### Where can I find my past projects? {#find-past-projects}
**There are two ways to find your past projects:**
**Method 1: Using the Task List**
1. Click the menu icon (two horizontal lines) in the top-left corner of the homepage or any Genspark project page
2. Search for your past projects using relevant keywords
**Method 2: Through Your Account**
1. Click the profile icon in the bottom-left corner of the sidebar
2. Click on your account name/email to enter your account page
3. Select "History" to view all your past projects sorted by last update time
4. Alternatively, directly visit https://www.genspark.ai/me and select "History"
### How do I delete my account? {#delete-account}
**Step-by-step process:**
1. Click Settings in the bottom left corner of the main page
2. Find the "Delete User" option and click it
3. Confirm deletion in the popup dialog and email
**Account deletion is PERMANENT and IRREVERSIBLE!** Once you confirm "Delete User", all your data will be permanently removed and cannot be recovered under any circumstances.
**This includes:**
- All conversation history
- Generated content (images, videos, text)
- Personal settings and preferences
- Unused credits
- Custom configurations
- Saved files and projects
**How to verify deletion:**
If you cannot log in with your email address after deletion, your account has been successfully removed from the system.
### What should I do if I notice suspicious account activity? {#suspicious-activity}
**If you notice unusual account activity:**
1. Immediately reset your password using the "Forgot Password" option
2. Log out from all devices if possible
3. [Submit a ticket](#support){.helpcenter-support-link} to contact support
4. Review recent activity: check for conversations or content you didn't create
**Signs of suspicious activity:**
- Conversations you don't remember having
- Password reset emails you didn't request
- Login notifications from unfamiliar locations
- Changes to your account settings you didn't make
# Membership Plans
Source: https://www.genspark.ai/helpcenter/membership-plans
## Plan Overview {#plan-overview data-section-type=overview}
Genspark membership plans give you credits for AI tasks, AI Drive storage, and access to the product's agents and models. This article explains what each plan includes, how Free Quota and credits work, and how to manage, change, or cancel your subscription.
Genspark offers these plans:
- **Free** — the default plan for every new account, with 100 credits per day and 1 GB of AI Drive storage.
- **Plus** — from 10,000 monthly credits, 50 GB of AI Drive storage, Free Quota, and commercial use of AI-generated content.
- **Pro** — everything in Plus, starting from 125,000 monthly credits, 1 TB of AI Drive storage, and a higher AI Chat limit.
- **Team** — a per-seat plan for organizations, with centralized admin and billing. See [Team & Enterprise Plans](https://www.genspark.ai/helpcenter/team-enterprise-plans).
- **Enterprise** — everything in Team plus contract-level governance, automatic user provisioning (SCIM), and support. See [Team & Enterprise Plans](https://www.genspark.ai/helpcenter/team-enterprise-plans).
The membership benefits described on this page are valid through June 30, 2027.
## Prices {#prices data-section-type=general}
All prices are in USD.
| Plan | Monthly | Annual |
| --- | --- | --- |
| Free | $0 | — |
| Plus | $24.99 / month | $19.99 / month, billed annually at $239.99 / year |
| Pro | $249.99 / month | $199.99 / month, billed annually at $2,399.99 / year |
| Team | $30 / seat / month | — |
| Enterprise | Custom | Custom |
## Plan Comparison {#plan-comparison data-section-type=general}
| | Free | Plus | Pro |
| --- | --- | --- | --- |
| Monthly credits | 100 / day, up to a lifetime limit | 10,000 | 125,000 |
| AI Drive storage | 1 GB | 50 GB | 1 TB |
| Free Quota (Standard mode) | — | Weekly balance | 10× the Plus balance |
| AI Chat & AI Image | Consumes daily credits | Core models at no credit cost; flagship models use credits | Core models at no credit cost; higher AI Chat limit; flagship models use credits |
| Speakly voice input | 4,000 words / week | No weekly word limit | No weekly word limit |
Team and Enterprise plans add seats, centralized admin and billing, usage analytics, connector management, and support options. See [Team & Enterprise Plans](https://www.genspark.ai/helpcenter/team-enterprise-plans).
## Free Plan {#free-plan data-section-type=general}
Every Genspark account starts on the Free plan. Free includes:
- 100 credits per day, up to a lifetime allowance of free credits
- 1 GB of AI Drive storage
- Limited access to AI models and agents
- Speakly voice input, 4,000 words per week
### Why did my daily credits stop? {#why-did-my-daily-credits-stop}
Free accounts receive a daily refill of 100 credits up to a lifetime allowance. Once your account has used that lifetime allowance, the daily refill stops — your current balance stays and keeps working until it runs out. If you see "You reached your lifetime limit of daily free credits," your account has reached that allowance. Upgrading to Plus or Pro replaces the daily Free credits with your plan's monthly allocation.
## Plus Benefits {#plus-benefits data-section-type=general}
Genspark Plus provides the following key benefits:
### Professional Workspace {#plus-professional-workspace}
- Full access to AI Slides, Genspark Code, and more agents
- 50 GB AI Drive storage for your projects
- Commercial use of all AI-generated content
### Credits {#plus-credits}
Plus offers multiple credit tiers, starting from 10,000 credits per month. Higher tiers provide more credits; you can choose your tier when subscribing or switch anytime.
### Free Quota {#plus-free-quota}
- Standard mode usage across agents — a weekly balance for text and image tasks in Standard mode across Super Agent, AI Slides, AI Sheets, AI Docs, Deep Research, Fact Check, and SecondBrain, at no credit cost. See [How Free Quota Works](#how-free-quota-works).
- AI Chat & AI Image — core models are included at no credit cost. When you reach the limit, you continue on your plan's monthly credits; flagship models use credits. See [About AI Chat & AI Image](#about-ai-chat-and-ai-image).
- AI Meeting Notes — meeting transcription and AI summaries included, up to 24 hours a day.
- Speakly — voice dictation included, zero credit cost.
## Pro Benefits {#pro-benefits data-section-type=general}
Genspark Pro includes all Plus benefits with professional-level upgrades:
### Professional Workspace {#pro-professional-workspace}
- Full access to AI Slides, Genspark Code, and more agents
- 1 TB AI Drive storage
- Commercial use of all AI-generated content
### Credits {#pro-credits}
Pro offers multiple credit tiers, starting from 125,000 credits per month. Higher tiers provide more credits; you can choose your tier when subscribing or switch anytime.
### Free Quota {#pro-free-quota}
- Standard mode usage across agents — a weekly balance 10× the Plus balance. See [How Free Quota Works](#how-free-quota-works).
- AI Chat & AI Image — the same core models as Plus, with a higher limit in AI Chat. See [About AI Chat & AI Image](#about-ai-chat-and-ai-image).
- AI Meeting Notes — meeting transcription and AI summaries included, up to 24 hours a day.
- Speakly — voice dictation included, zero credit cost.
## How Free Quota Works {#how-free-quota-works data-section-type=general}
Free Quota is a weekly balance included with Plus and Pro. It covers text and image tasks in Standard mode across Super Agent, AI Slides, AI Sheets, AI Docs, Deep Research, Fact Check, and SecondBrain, at no credit cost.
### The weekly window starts when you do {#the-weekly-window-starts-when-you-do}
Your 7-day window begins with your first message and resets 7 days later.
### What Free Quota covers {#what-free-quota-covers}
Standard mode in the agents above; in media agents, planning is covered and the generated video or audio uses credits. Coverage expands over time — more products and modes will be added.
### How you know Free Quota is active {#how-you-know-free-quota-is-active}
A blue bar above the composer reads "Using free quota at no credit cost" whenever a task is covered.
### Checking your remaining Free Quota {#checking-your-remaining-free-quota}
Your profile menu shows your current credits and remaining Free Quota. For details, open your usage page.
### When Free Quota runs out {#when-free-quota-runs-out}
Tasks switch to your credits automatically and keep running; the balance takes over again at the next reset. Turn on Ask First on your usage page if you prefer to confirm before credits are used.
### On mobile {#free-quota-on-mobile}
Members who subscribed through the App Store or Google Play have the same Free Quota, shared with the web. The app has no blue notice bar: Free Quota automatically applies when you run a covered task in Standard mode. To check your remaining Free Quota, open the usage page.
## About AI Chat & AI Image {#about-ai-chat-and-ai-image data-section-type=general}
### AI Chat {#about-ai-chat}
Core models are included at no credit cost; the model selector lists them under "Core models" with a "No credit cost" tag. When you reach the limit, there's no waiting: you continue on your monthly credits right away. Flagship models sit under "Flagship models" with a "Credits" tag and use credits.
### AI Image {#about-ai-image}
Models tagged "No credit cost" are included at no credit cost, for example GPT Image at standard quality. When you reach the limit, there's no waiting: you continue on your monthly credits right away. Image models without the "No credit cost" tag use credits. Some of GPT Image's higher-quality and larger-size options use credits and are marked "Credits" individually.
These are examples only. The models available to you are shown in the product, and the list updates as new models launch. Credit-billed models stay available to every member.
## Members Who Subscribed Before September 18, 2026 {#members-who-subscribed-before-september-18-2026 data-section-type=general}
If your Plus or Pro subscription started before September 18, 2026 and has remained active without interruption, this section applies to you. It covers both monthly and annual plans.
### Free Quota is added, and your plan just got bigger {#free-quota-is-added-and-your-plan-just-got-bigger}
Starting September 18, 2026, you receive Free Quota just like new members: a weekly balance for Standard-mode tasks across Super Agent, AI Slides, AI Sheets, AI Docs, and more, at no credit cost. See [How Free Quota Works](#how-free-quota-works).
### Everything you already have stays {#everything-you-already-have-stays}
The credits, models, professional workspace, commercial use rights, AI Meeting Notes, and Speakly included in your plan remain exactly as they are. AI Chat and AI Image also keep your current unlimited usage, within the same 5-hour session window described in [About AI Chat & AI Image](#about-ai-chat-and-ai-image), unchanged through December 31, 2026.
### From January 1, 2027 {#from-january-1-2027}
AI Chat and AI Image benefits for all Plus and Pro members follow the plan terms in effect at that time: core models are included, and once you reach the usage cap you continue with your monthly credits; flagship models use credits. See [About AI Chat & AI Image](#about-ai-chat-and-ai-image).
### What counts as uninterrupted {#what-counts-as-uninterrupted}
Your subscription counts as uninterrupted as long as it has not expired. Upgrading (for example, Plus to Pro), switching between monthly and annual billing, or changing your credit tier does not affect this arrangement. If your subscription expires, or you cancel and later resubscribe, the benefits in effect at the time you resubscribe apply.
## Subscription Management {#subscription-management data-section-type=general}
### How to Subscribe {#how-to-subscribe}
Follow these simple steps to subscribe:
1. Go to the plan page ([click here](/me?open_pricing=help_center)), choose your preferred plan and billing cycle, and click **Upgrade**:
- Monthly plan — billed each month
- Annual plan — billed as a lump sum for the full year upfront (saves about 20% compared to monthly); credits are issued on a monthly basis throughout the year
2. On the Stripe payment page, enter your payment details securely.
3. Confirm your subscription.
4. You'll be redirected back to Genspark.
5. An invoice email will be sent to your payment email shortly.
### Subscribing on mobile {#subscribing-on-mobile}
You can also subscribe directly from the Genspark iOS or Android app. Tap your profile icon, choose a plan, and complete your purchase through the App Store (iOS) or Google Play (Android). See [Mobile Subscriptions](#mobile-subscriptions) for details.
### Changing Plans {#changing-plans}
You can modify your subscription plan at any time from the plan page ([click here](/me?open_pricing=help_center)), including switching between Free, Plus, and Pro plans or adjusting other aspects of your current plan. Manage Plan opens the pricing panel in place, where you can review plans, change tiers, or cancel.
#### Upgrading your plan {#upgrading-your-plan}
- Takes effect immediately
- You are charged a prorated amount for the remainder of your current billing period
- Supplemental credits (the difference between your new and old plan's monthly credits) are added to your account right away, valid through the end of your current billing period
- Starting from your next billing cycle, the full credit allocation of your new plan applies
**Example:** If your Plus monthly plan renews on March 1 and you upgrade to Pro on March 15, you'll be charged a prorated amount for the remaining ~16 days of March. Your account will also immediately receive the supplemental credits (the difference between Pro and Plus monthly credit allocations), valid through March 31. Starting April 1, you'll receive the full Pro credit allocation on each renewal.
#### Downgrading your plan {#downgrading-your-plan}
- Your current plan remains active through the end of your current billing period
- The new plan takes effect on your next renewal date
- Credits are not carried over; your credit allocation resets to the new plan's amount at the start of the next billing cycle
**Example:** If your Pro plan renews on April 1 and you request a downgrade to Plus on March 15, your Pro benefits remain fully active through March 31. On April 1, your plan switches to Plus and your credit balance resets to the Plus monthly allocation — unused Pro credits are not carried over.
### Canceling Your Subscription {#canceling-subscription}
You can cancel your subscription at any time. Open Settings, go to Billing, and find the Cancellation card. Select Cancel Subscription.
- After canceling, your current plan remains active through the end of your billing period — you won't lose access right away
- Once the billing period ends, your account automatically reverts to the Free plan (100 daily credits, until your lifetime free-credit allowance is used up — see the [Credits Guide](https://www.genspark.ai/helpcenter/credits-guide))
- If you change your mind, you can reactivate your subscription at any time before the period ends to continue without interruption
- If your payment is past due, cancellation takes effect immediately instead of at the end of the billing period.
#### If the billing portal won't open {#if-the-billing-portal-wont-open}
If you subscribed through the App Store or Google Play, manage or cancel the subscription from your device's subscription settings. The Genspark website can't manage an app-store subscription.
If you subscribed on the web and the billing portal won't open, contact our support team by clicking Support in the menu and include the email on your Genspark account. Support can help you cancel or change the subscription.
#### Subscribed via iPhone or Android {#subscribed-via-iphone-or-android}
If you subscribed through the App Store or Google Play, you must cancel through your device's subscription settings — canceling on the Genspark website will not affect a mobile subscription. See [Mobile Subscriptions](#mobile-subscriptions) for details.
## Membership Transfer {#membership-transfer data-section-type=general}
Subscriptions are tied to the Genspark account you were logged into at the time of purchase and cannot be transferred to another account.
If you need to switch to a different account, you can cancel your subscription on the current account and purchase a new subscription on the account you'd like to use.
## Mobile Subscriptions {#mobile-subscriptions data-section-type=general}
You can subscribe to Genspark Plus or Pro directly from the Genspark mobile app:
- **iOS** — Purchase through the App Store's in-app purchase dialog
- **Android** — Purchase through Google Play (payment method may vary by region)
Important notes about mobile subscriptions:
- Mobile subscriptions are managed separately from web subscriptions — you cannot manage a mobile-purchased subscription from the web, and vice versa
- You can only have one active subscription platform at a time. If you already have a web subscription and try to subscribe on mobile (or vice versa), the app will notify you of the conflict
- To manage or cancel a mobile subscription, use your device's subscription settings (App Store settings on iOS, or Google Play subscriptions on Android)
- Payment details and invoices for mobile purchases are handled by the respective app store, not through Stripe or the Genspark billing portal
## Genspark Claw {#genspark-claw data-section-type=general}
Genspark Claw runs on Genspark Cloud Computer for deployment. To use Genspark Claw, you'll need an active Genspark Cloud Computer subscription.
Genspark Cloud Computer currently offers three subscription tiers:
| Tier | Specs |
| --- | --- |
| Lite | 2 vCPU / 4GB RAM / 64GB Storage |
| Standard | 2 vCPU / 8GB RAM / 64GB Storage |
| Powerful | 4 vCPU / 16GB RAM / 128GB Storage |
### Credits usage {#cloud-computer-credits-usage}
Cloud Computer provides the runtime environment, and some Claw actions may use Genspark credits. Credits are shared and managed in one place across Genspark. If you need more credits, you can upgrade to a Plus or Pro plan. Credit usage varies based on the task type, the agent you use, and the specific features involved. To get more credits, you can upgrade your membership or increase your credit tier if you're already a Plus or Pro subscriber.
### Canceling Cloud Computer {#canceling-cloud-computer}
You can cancel your Genspark Cloud Computer subscription at any time. Cancellation takes effect at the end of your current subscription period. Once it takes effect, the Cloud Computer is reclaimed and all data stored on it is permanently erased and cannot be recovered, so back up any critical data before cancelling.
## Team & Enterprise Plans {#team-and-enterprise-plans data-section-type=general}
Team and Enterprise plans add seats, centralized admin and billing, usage analytics, connector management, and support options for organizations.
Team includes 12,000 credits per seat per month. Enterprise credits are set per Order Form (custom), so a seat's allocation can be tailored to your contract.
For full details, see [Team & Enterprise Plans](https://www.genspark.ai/helpcenter/team-enterprise-plans).
## Payment & Billing {#payment-billing data-section-type=general}
The payment and billing information below applies to web subscriptions processed through Stripe. If you subscribed through the iOS App Store or Google Play, your payment details, invoices, and billing are managed by the respective app store. See [Mobile Subscriptions](#mobile-subscriptions) for details.
### Payment Security {#payment-security}
We use Stripe to handle transactions. Your raw payment information is encrypted and never stored on our servers. Industry-standard security protocols protect your data throughout the payment process.
### Finding Your Invoice {#finding-invoice}
Your payment receipt will be automatically sent to the billing email address you provided at the time of purchase. If you do not see it in your inbox, check your spam or promotions folder as well.
You can access your billing portal at any time from the plan page ([click here](/me?open_pricing=help_center)). Open Manage Plan and choose Edit Billing / Payment Method to view past invoices and manage your payment details.
### Update Payment Method {#update-payment-method}
To update your payment method, go to the plan page ([click here](/me?open_pricing=help_center)). Open Manage Plan and choose Edit Billing / Payment Method — this will open your billing portal where you can add or change your card.
### Understanding the Final Price {#final-price}
The displayed price reflects the base membership rate. Genspark does not charge any fees beyond this amount. However, your final payment may be slightly higher due to charges from third parties:
- **Sales tax (VAT / GST)** — Collected by local tax authorities based on your region; Genspark has no control over this amount
- **Currency conversion fees** — Applied by your bank when paying in a currency other than USD
- **Foreign transaction fees** — Charged by your bank or card issuer for cross-border transactions
These additional charges, if any, are collected by your bank or tax authority, not by Genspark.
### When Payment Fails {#payment-fails}
We don't charge until the payment process succeeds. You can try again with the same or different payment method. No charges will appear on your account until successful payment is confirmed.
### I Paid But Didn't Get Plus/Pro {#paid-no-upgrade}
This should not happen. If it does, please try the following first:
**Common cause:** You may have entered a different email on the Stripe payment page. That email is only used for receiving invoices — your subscription is linked to the Genspark account you were logged into when you started checkout. Try logging in with your original Genspark registration email, then refresh the page and check your plan status again.
If the issue persists, please contact our support team by clicking Support in the menu, and provide the following details:
- Your name
- Your Genspark login email
- The email used for payment (if different from your login email)
- Your invoice number, or a screenshot of the invoice email
Do NOT include sensitive information such as:
- Credit card numbers
- Home address
- Other personal financial details
### Invoice and billing support {#invoice-and-billing-support}
For invoice questions, contact our support team by clicking Support in the menu and include your invoice number. To view past invoices, open the billing portal from the plan page (see [Finding Your Invoice](#finding-invoice)).
Please do not include sensitive information such as credit card numbers, home address, or other personal financial details in your request.
## Refunds & Cancellations {#refunds-and-cancellations data-section-type=general}
### Refund Policy {#refund-policy}
Our refund policy depends on your subscription type, local law, and account status. We can process only one refund per customer — once a refund request has been approved, a second request can't be approved. This applies to Plus/Pro memberships and to Genspark Cloud Computer subscriptions.
The refund window is measured from the charge:
- **EU, UK, and Turkey:** 14 days for both monthly and annual subscriptions. Please state in your request that you are requesting a refund from the EU, UK, or Turkey.
- **Korea:** 7 days for both monthly and annual subscriptions.
- **All other customers:** 72 hours for both monthly and annual subscriptions.
Within the eligible window, a refund is available if you have used less than 50% of the subscription or quota. If you have used 50% or more, we may decline the refund and offer credits instead.
If a refund is approved, the subscription is cancelled at the same time.
Refund requests are normally accepted only within the applicable window; requests submitted after the window are reviewed case by case.
Team and Enterprise refund requests are reviewed by our team. Enterprise coverage continues through the contracted end date.
For purchases made through the App Store or Google Play, refunds are handled by the app store — for iOS, use [reportaproblem.apple.com](https://reportaproblem.apple.com/).
### Refunds after a renewal {#refunds-after-a-renewal}
You can request a refund for an automatic renewal charge, not only for the initial purchase. Each charge starts its own refund window, so a renewal is refundable on the same terms as a first purchase.
### Requesting a Refund {#requesting-refund}
To request a refund within the eligible period, submit a ticket. The ticket is automatically linked to the Genspark account you're logged into, so please make sure you're logged into the account that made the purchase before submitting.
In your ticket, please include:
- The email used for payment (if different from your login email)
- Your invoice number, or a screenshot of the invoice email — you can paste or drag images directly into the ticket form (see [Finding Your Invoice](#finding-invoice))
- Reasons for requesting a refund
**Duplicate request notice:** Please submit only one ticket per refund request. Multiple tickets about the same request may delay processing. All requests are handled in the order received.
**Processing time:** Approved refunds are processed the same day. Depending on your bank, the money arrives within 5–10 business days.
## Fair Use Limit {#fair-use-limit data-section-type=general}
To keep Genspark fast, fair, and reliable for everyone, we have guardrails in place to prevent abuse.
Prohibited activities include:
- Automatically or programmatically extracting data in bulk
- Sharing your account credentials or making your account available to others
- Reselling access or using Genspark to power third-party services
## FAQ {#faq data-section-type=faq}
### Why did a task use credits when I still have Free Quota? {#why-did-a-task-use-credits-when-i-still-have-free-quota}
The task wasn't in a covered scope yet — for example, it ran in Ultra mode, or generated video or audio. Covered tasks show the blue "Using free quota" bar above the composer, and coverage expands over time.
### I subscribed before September 18, 2026. Does anything change for me? {#i-subscribed-before-september-18-2026-does-anything-change-for-me}
If your Plus or Pro subscription started before September 18, 2026 and stays active without interruption, your current unlimited AI Chat and AI Image usage continues unchanged through December 31, 2026 — within the same 5-hour session window described in [About AI Chat & AI Image](#about-ai-chat-and-ai-image) — and you receive Free Quota on top. This applies to monthly and annual plans alike. If your plan lapses and you re-subscribe, the benefits in effect at that time apply. From January 1, 2027, AI Chat and AI Image benefits for all Plus and Pro members follow the plan benefits in effect at that time.
### Can I switch between monthly and annual billing? {#switch-billing-interval}
Yes. You can switch directly from the plan page ([click here](/me?open_pricing=help_center)) — no need to cancel and resubscribe. Choose the billing interval you'd like to switch to and confirm.
- **Monthly → annual:** The switch takes effect immediately. You'll be charged a prorated amount for the price difference.
- **Annual → monthly:** The switch is scheduled for your next billing cycle. You'll continue on the annual plan until the current period ends, then automatically switch to monthly billing.
### Can I upgrade from Plus to Pro? {#upgrade-plus-to-pro}
Yes, you can upgrade at any time. The upgrade takes effect immediately, and you'll be charged a prorated amount for the price difference covering the rest of your current billing period.
### What happens to my credits when I upgrade? {#credits-on-upgrade}
It depends on whether your billing interval changes:
- **Same billing interval (e.g., Plus Monthly → Pro Monthly):** You keep your existing credits, and additional credits equal to the difference between the two plans are added immediately for the remainder of your current billing period.
- **Billing interval change (e.g., Plus Monthly → Pro Annual):** Your unused credits from the previous plan are revoked, and a full set of credits under the new plan is issued starting immediately.
### What happens when I cancel? {#cancellation-effects}
Your subscription is not terminated immediately. Your current plan and credits remain fully active until the end of the billing period you've already paid for. After that, your account reverts to the Free plan (100 daily credits until your lifetime free-credit allowance is used up). You can reactivate at any time before the period ends to continue without interruption.
### Can I pause my subscription? {#can-i-pause-my-subscription}
Pausing a subscription isn't available. You can cancel and resubscribe later, but the benefits in effect at the time you resubscribe apply.
### I got a new device. How do I restore my purchase? {#i-got-a-new-device-how-do-i-restore-my-purchase}
Sign in with the same Genspark account you used to subscribe — your subscription is tied to your account, not your device. For a purchase made through the App Store or Google Play, open the app's subscription or restore-purchases option and use the same app-store account.
### How many GenTeam agents can I run on my plan? {#how-many-genteam-agents-can-i-run-on-my-plan}
| Limit | Free | Plus, Pro, Team, Enterprise |
| --- | --- | --- |
| Local agents (per computer) | 3 | 100 |
| Agents | 100 | 200 |
| Channels | 100 | 10,000 |
| Local agents | 100 | Unlimited |
Plus, Pro, Team, and Enterprise share the same agent limits.
### When do credits post after payment? {#when-do-credits-post-after-payment}
Credits are added when your payment is confirmed. If you pay annually, credits are issued monthly across the year rather than all at once. If your plan is active but your credits haven't appeared, contact support with your invoice number.
### Why didn't my promo code apply? {#why-didnt-my-promo-code-apply}
Promo codes are validated at checkout and may be limited by account, region, or expiry date. If a code doesn't apply, check that it's entered exactly as written and hasn't expired. We can't confirm why a specific code didn't work from the help center — contact support if you need help with a code.
### How do I get more credits? {#get-more-credits}
The most straightforward way is to upgrade your plan. Both Plus and Pro offer multiple credit tiers — upgrading to a higher tier gives you more monthly credits. You can view available tiers and upgrade from the plan page ([click here](/me?open_pricing=help_center)).
### Can I use AI-generated content for commercial purposes? {#commercial-use}
Yes. All Genspark Plus and Pro subscribers are granted commercial use rights for AI-generated content created during their active subscription period. This covers content generated across Genspark's AI tools, including images, video, audio, and other outputs.
Commercial use rights are guaranteed through June 30, 2027 as part of your Plus or Pro membership benefits. For content created after that date, please refer to the latest terms on the plan page.
# Team & Enterprise Plans
Source: https://www.genspark.ai/helpcenter/team-enterprise-plans
## What is Team and Enterprise Plan? {#overview data-section-type=overview}
**Team** and **Enterprise** are Genspark's two group subscription tiers — they let an entire organization use Genspark together under a single subscription, with centralized billing and admin controls. Both plans share the same dashboard, the same admin tools, and the same per-member experience; they differ mainly in **scale, contract structure, support level, and customizable governance**.
Both plans share the core promise: every member gets an independent Genspark workspace, the team is automatically opted out of model training, admins control billing / seats / connectors / SSO, and admins **cannot** view individual members' project content.
### Team Plan benefits {#team-benefits}
Team is self-serve and built for groups of **2–150 people** that want to start collaborating right away. Each seat includes:
- **12,000 credits per month** — individually allocated, reset each billing cycle, do not roll over
- **60 GB AI Drive storage** — per-seat, private to each member
- **Full access to all Genspark agents** — AI Chat, AI Slides, AI Docs, AI Sheets, AI Image, AI Video, Genspark Code, Super Agent, and more
- **All latest video generation models** — Gemini Veo, Kling, Runway, Seedance, MiniMax Hailuo, PixVerse, Wan, Hunyuan, and more
- **All latest audio generation models** — Gemini, ElevenLabs, MiniMax, Mureka, Lyria2, and more
- **Unlimited zero-credit usage of top-tier chat & image models** in AI Chat Agent and AI Image Agent — access the latest SOTA text and image models at zero credit cost, *valid until December 31, 2026*. The lineup is continuously updated; see the live list on the [pricing page](https://www.genspark.ai/team_pricing)
- **Commercial use rights** for all AI-generated content — *valid through December 31, 2026*
- **Self-serve monthly billing** via credit card (Stripe); cancel anytime
Pricing: **$30 / seat / month**.
### Enterprise Plan benefits {#enterprise-benefits}
Enterprise is contracted and built for organizations with **151+ users** that need tailored terms, dedicated support, and advanced governance. Each seat includes:
- **25,000 credits per month** — individually allocated; custom credit pools available per Order Form
- **Custom AI Drive storage capacity** — negotiated in your Order Form
- **Full access to all Genspark agents** and **all video / audio / chat / image models** — standard credit pricing applies (no zero-credit promo); typically offset by the higher per-seat allocation and Order Form–negotiated terms
- **Commercial use rights** confirmed in your Order Form
- **Custom contract** — Order Form + Master Subscription Agreement (typically 36-month initial term, 12-month auto-renewals)
- **Dedicated support** — Customer Success Manager, 4-hour critical-issue response, 24/7 coverage for critical issues, 99.9% uptime SLA with service credits (per Order Form)
- **Advanced governance** — agent-level permissions, AI model restrictions, organization-wide credit pool and usage caps, login and session logs, and a per-member breakdown of credit usage
- **Compliance & security** — custom DPA, configurable data residency (US / EU / APAC), dedicated VPC option, custom compliance addendums (ITAR/EAR, BAA, industry-specific), 24-hour security incident notification with dedicated security contact
- **Wire transfer / ACH / invoice** billing — Net 30; Purchase Orders accepted in lieu of an Order Form
Pricing: **contact sales**. Enterprise pricing is negotiated based on factors such as usage and any custom configuration (e.g., per-seat credit allocation), and is locked in when the Order Form is signed.
For full per-dimension detail, see [Full feature comparison](#feature-comparison) near the bottom.
## Quick Start {#quick-start data-section-type=overview}
### Choose your plan: Team or Enterprise {#choose-plan}
- **2–150 people, want to start today, fine with monthly self-serve billing** → **Team Plan**. Sign up at [genspark.ai/team_pricing](https://www.genspark.ai/team_pricing).
- **151+ people, OR you need a custom contract / DPA / data residency / dedicated support / advanced governance** → **Enterprise Plan**. Contact `enterprise@genspark.ai` to start the conversation.
If you're between the two — for example, a 100-person team that needs custom DPA terms or wire transfer billing — Enterprise can still be the right fit. Talk to sales.
For a complete dimension-by-dimension comparison covering pricing, features, governance, compliance, data privacy, retention, security architecture, and support, see [Full feature comparison](#feature-comparison) near the bottom of this page.
### Team Plan: Self-serve sign-up {#team-signup}
1. Go to [genspark.ai/team_pricing](https://www.genspark.ai/team_pricing) and sign in (or create an account if you don't have one)
2. Pick the number of seats you need (minimum **2**, up to **150**) — the displayed monthly cost updates automatically
3. Click **Get Team Plan** to continue to Stripe checkout
4. Enter your payment details and confirm
5. After checkout you'll be redirected to the **Team Plan Management** page (also reachable any time from the pricing page → **Manage Team Subscription**)
The user who completes checkout becomes the first **Admin** of the team.
### Enterprise Plan: Sales-assisted onboarding {#enterprise-signup}
1. Email `enterprise@genspark.ai` with your organization name, expected seat count, and any specific requirements (compliance, data residency, support tier, etc.)
2. Genspark sales will follow up to discuss your requirements and prepare an **Order Form** with custom pricing and terms
3. Sign the Order Form (or submit a Purchase Order if accepted in lieu)
4. Genspark provisions your organization and grants the first **Admin** access to the management dashboard
5. Your Customer Success Manager schedules onboarding and reviews any deployment-specific configuration (SSO, residency, custom DPA, etc.)
### Invite your team {#invite-team}
After your team is provisioned, open the **Members** tab in the management dashboard and invite members by email or via the public invite link. See [Invite members](#invite-members) under Managing Your Team for the detailed flow.
## Managing Your Team {#managing-team data-section-type=general}
This section is for organization admins. Regular members can skip ahead to [Member Experience](#member-experience).
### Dashboard overview {#dashboard-overview}
Both Team and Enterprise admins use the same management page. Reach it any time from the pricing page → **Manage Team Subscription**, or directly at `/payment/team_admin`.
Only members with the **Admin** role can open this page. Regular members do not see the dashboard.
The left sidebar (**Organization Payment**) gives admins access to:
- **Billing** — Subscription details, seats, credits, payment method, invoices
- **Members** — Invite or remove team members
- **Connectors** — Choose which third-party connectors are available to your team
- **User analytics** — Team-wide credit usage charts (**Last 30 Days / All Time**), per-member usage and CSV export under the **Member Stats** sub-tab; Enterprise admins also see **Usage Logs** and **Login History** sub-tabs
- **Identity & provisioning** — SAML SSO configuration
- **API Keys** — View and revoke API keys created by team members
### Member roles {#member-roles}
- **Admin** — Full access to the dashboard. A team can have multiple admins; promote any member to admin from the **Members** tab.
- **Member** — Uses Genspark normally with their own credits and storage. Cannot see the admin dashboard.
On Team Plan, every admin counts as a paid seat.
### Members {#members}
Open the **Members** tab. It has two sub-tabs: **Members** and **Invites**. Per-member usage stats, CSV export, and credit-pack assignment live under **User analytics → Member Stats** — see [Track usage and assign extra credits](#track-usage) below.
#### Invite members {#invite-members}
The invite flow is identical on both plans. From the management page, open the **Members** tab and click **Invite Member**:
- Enter one or more email addresses (one per line), or upload a CSV
- The team's **Public invite link** is also shown in the invite dialog — anyone with the link can join while seats are available
- Click **Reset link** to invalidate the current invite link if you want to stop new joiners using the old URL
When an invitee accepts, they consume one seat. Pending invites are visible under the **Invites** sub-tab. You cannot invite more members than you have available seats — to add seats, see [Billing & seats](#billing-seats) below.
#### Remove a member {#remove-member}
On the Members sub-tab, find the member's row and click the actions menu → **Remove Member**.
- The member loses access immediately
- Their seat becomes available so you can invite a replacement
- **Unused credits are not transferred** to other members or back to the admin — this is by design on both plans
#### Track usage and assign extra credits {#track-usage}
In the sidebar, open **User analytics** → switch to the **Member Stats** sub-tab:
- Filter members by email
- See each member's current balance, credit-pack claims, and usage breakdown for the last 7 / 30 days or all time
- Export the full member-stats table as CSV
To give a specific member extra credits, click **Assign credit pack** on their row and select the number of packs to assign. This requires available packs in your team's pool — purchase more from **Billing → Add Credits** if needed.
### Billing & seats {#billing-seats}
Open **Billing** in the sidebar to see your subscription details: plan, active seats, monthly cost (pre-tax), and next payment date.
#### Team Plan billing {#team-billing}
- **Payment** — Credit card only, processed via Stripe. Bank transfer / ACH / invoice are not available on Team.
- **Renewal** — Monthly auto-renewal. Cancel anytime; coverage continues to the end of the paid period.
- **Manage Seats** — Click **Manage Seats** to add or remove seats. The dialog previews the cost impact before you confirm.
- **Increasing seats** takes effect immediately. You're charged a **prorated** amount for the new seats from today through your next renewal date
- **Decreasing seats** takes effect at the **start of your next billing cycle**. Until then, current seats remain active and billable
- You cannot decrease below 2 seats
- **Add Credits** — Purchase **Credits Packs** when your team is running low (see [Credits Packs](#credits-packs))
- **Manage Payment Method** — Click to update or replace the card on file
- **Invoices** — All invoices are listed under **Invoices** in the Billing tab; click any invoice to download
#### Enterprise Plan billing {#enterprise-billing}
- **Payment** — Wire transfer, ACH, or invoice; Net 30 payment terms.
- **Purchase orders** — POs accepted in lieu of an Order Form, subject to standard terms.
- **Initial term** — Typically 36 months, with 12-month auto-renewals; specific terms are defined in your Order Form.
- **Termination** — 60 days written notice required.
- **Billing dispute window** — 60 days from the first statement.
- **Seat changes** — Adjusted via your Order Form (or amendment); your CSM coordinates seat additions, true-ups, and adjustments.
- **Custom credits / pools** — Enterprise can be configured with shared credit pools, dedicated credit packs, or custom allocations per your Order Form.
- **Invoices and payment methods** — Managed through the contracts your CSM sets up; the dashboard's standard "Manage Payment Method" flow may not apply for invoice-billed Enterprise accounts.
#### Credits Packs {#credits-packs}
On Team Plan, use **Add Credits** under Billing to purchase **Credits Packs** when team members are running through their per-seat allocation.
- **10,000 credits per pack**
- Once a member **claims** a pack, those credits are added to their account; the pack's specific validity window is shown at purchase time
- Members claim packs in one of two ways:
- **Self-claim** — when a member runs out of their monthly credits, a popup lets them claim an available pack with one click. Both the member and all admins are notified by email
- **Admin assignment** — admins can pre-assign packs to specific members from **User analytics → Member Stats** (**Assign credit pack**)
- Unclaimed packs stay in the team's pool and can be claimed or assigned later
Enterprise customers can use the same packs flow, or negotiate alternative arrangements (shared credit pools, custom expiration windows, etc.) in their Order Form.
#### Centralized payment for the whole team {#centralized-payment}
A team can be paid for from a single payment source — for example, an admin pays with a company card on Team, or accounts payable processes a wire on Enterprise. Members do not need their own payment methods to use the team subscription.
If a member previously had a personal Plus or Pro subscription, they should **cancel that personal subscription before joining the team** to avoid being billed twice for overlapping plans. (If both subscriptions stay active under the same email, you'll be charged for both, and you'll receive each plan's monthly credit allocation as a separate credit grant on the same account — there is no discount or merging of the two subscriptions, though the credits do pool into a single spendable balance.)
### Connectors {#connectors}
In **Connectors**, admins control which third-party connectors (e.g., Google Drive, Notion, Slack) are available to organization members.
- Select the connectors you want enabled for the whole team and click **Save Changes**
- Settings apply org-wide on both plans — currently you cannot enable a connector for some members and disable it for others
- The dialog shows how many members will be affected by your change
### SSO setup {#sso-setup}
In **Identity & provisioning**, configure SAML 2.0 single sign-on so members can log in through your company's identity provider. Both Team and Enterprise plans support:
- Microsoft Entra ID (Azure AD)
- Google Workspace
- Okta
- Generic SAML 2.0
Enterprise customers may request additional provider integrations or custom claim mappings via the Order Form.
**Setup steps:**
1. Choose your **Provider Type**
2. Enter your **Email Domain** (e.g., `yourcompany.com`)
3. Fill in the SAML connection details from your IdP:
- **SSO URL** — the SAML 2.0 endpoint
- **Entity ID (Issuer)** — the unique identifier for your IdP
- **X.509 Certificate** — the public certificate from your IdP
- **Logout URL** *(required)* — the SAML logout endpoint
4. Copy the **ACS URL (Reply URL)** shown on the page and configure it in your IdP as the Assertion Consumer Service URL
5. Toggle **Enable SSO** and click **Save Configuration**
**Optional — require SSO for all members:** Toggling **Require SSO for all members** blocks every other login method (password, Google, etc.) for users in your domain. **Before enabling this**, open an incognito window and complete a full SSO login through your IdP to confirm the configuration works. If SSO is misconfigured, enforcing it can lock out you and every member of the organization — there is no in-app fallback.
### Enterprise-only admin features {#enterprise-admin}
When your organization is on Enterprise, the following capabilities are unlocked in the same dashboard:
- **Usage Logs** — Under **User analytics → Usage Logs**, export per-member credit usage as CSV with a selectable date range
- **Login History** — Under **User analytics → Login History**, review when and from where members signed in, for security audit and incident investigation
- **Agent-level permissions** — Admins can restrict members from using specific features
If your team needs any of these and you're currently on Team Plan, reach out to `enterprise@genspark.ai`.
## Member Experience {#member-experience data-section-type=general}
Once a member accepts an invite — same on both plans:
- They use Genspark just like any individual user — same agents, same UI
- They have **independent access**: they don't need any other member to be logged in, and their projects, files, and credits are tied to their own account
- They get their own dedicated credits (12,000/month on Team, 25,000/month on Enterprise) and storage allocation (60 GB on Team, custom on Enterprise)
- They **cannot access the admin dashboard** and cannot see other members' projects, files, or usage
When a member runs out of credits, if their team has unclaimed Credits Packs available, a popup will offer one-click claim. Otherwise the member can ask an admin to either purchase more packs or assign one directly to them.
## Security & Privacy {#security-privacy data-section-type=general}
Both plans include:
- **No training on your data** — Accounts are automatically opted out of model training. Per Genspark's Data Processing Agreement (DPA), your prompts and content are never used to train Genspark models, and our subprocessors (OpenAI, Anthropic, Google) are bound by no-training agreements with Zero Data Retention (ZDR) where applicable
- **Encryption** — Data is encrypted at rest (AES-256-GCM) and in transit (TLS 1.3 preferred, TLS 1.2+ supported)
- **Compliance** — SOC 2 Type II certified and ISO 27001:2022 certified. Trust Center: [app.vanta.com/mainfunc.ai/trust](https://app.vanta.com/mainfunc.ai/trust/9mmidu6lo2p5p3z7h1etc){rel="nofollow noopener noreferrer"}
- **Data ownership** — Customer retains all rights to customer data; Genspark retains all rights to its service / software IP
- **Admin visibility** — Admins can see seat-level credit usage and account activity, but **cannot view individual project history or content**
- **Account data deletion** — Account data is deleted 30 days after termination (Enterprise: extendable by request)
- **Backup retention** — 90 days post-deletion
### Enterprise additions {#enterprise-security}
Enterprise customers can additionally negotiate:
- **Custom DPA** tailored to compliance requirements
- **Configurable data residency** — US, EU, or APAC (vs. US-only default for Team)
- **Dedicated VPC option** for network isolation (vs. shared infrastructure on Team)
- **Custom compliance addendums** — ITAR/EAR, BAA, and other industry-specific terms
- **Customer-specific penetration test reports** in addition to those available in the Trust Portal
- **24-hour security incident notification** with a dedicated security contact (vs. standard notification on Team)
- **Continuous vulnerability scanning** with customer notification on findings
- **Certified deletion** upon request and **legal hold** capability for litigation support
If you need a SOC 2 Type II report or detailed audit reports, this is available to Enterprise customers under NDA — contact your CSM or `enterprise@genspark.ai`.
## Support & SLA {#support-sla data-section-type=general}
| | **Team Plan** | **Enterprise Plan** |
|---|---|---|
| Channel | Email support | Dedicated Customer Success Manager + priority queue |
| Critical response | 24 hours | 4 hours |
| High response | 48 hours | 8 hours |
| Hours | Business hours (M–F) | 24/7 for critical issues |
| SLA uptime | Not specified in Team MSA | 99.9% |
| Service credits | Not available | Service credits issued for downtime |
| Onboarding | Self-service documentation | Dedicated implementation team |
| Quarterly Business Reviews | No | Yes |
## Cancellation & Refunds {#cancellation-refunds data-section-type=general}
### Team Plan {#team-cancellation}
- Cancel any time from **Billing → Manage Payment Method** (your billing portal)
- Your team retains access through the end of the current billing period; no further charges will be made after cancellation
- Once the period ends, all member seats revert to the Free plan and admin features are turned off
- Team Plan subscriptions are **non-refundable** for partially used billing periods — credits already issued for the cycle are not credited back
### Enterprise Plan {#enterprise-cancellation}
- Termination requires **60 days written notice**
- Subscriptions are **non-refundable**; coverage continues through the contracted end date
- Account data is deleted 30 days after termination (extendable by request)
- Legal hold is available if your organization needs to preserve data for litigation
If you have a billing question or need to dispute a charge, contact our support team via the **Support** link in your Genspark account, or — for Enterprise — reach out to your CSM directly.
## Full feature comparison {#feature-comparison data-section-type=general}
This section is a comprehensive side-by-side reference for procurement, IT, and security reviewers. Earlier sections already cover the most common questions in detail; use this comparison when you need to walk through every dimension. Where a contractual number is shown, it reflects the standard Team Plan terms or the typical Enterprise starting position — Enterprise specifics may be adjusted in your Order Form. For binding language, refer to your executed agreement.
### 1. Pricing & Billing {#fc-pricing}
| Dimension | Team Plan | Enterprise Plan |
|---|---|---|
| Cost per seat | $30 / month | Contact sales (priced by usage and custom terms) |
| Monthly credits per seat | 12,000 | 25,000 |
| Minimum seats | 2 | 151 |
| Maximum seats | 150 | Unlimited |
| Contract structure | Month-to-month subscription | Custom contract — Order Form + Master Subscription Agreement |
| Initial term | Rolling monthly | Typically 36 months (per Order Form) |
| Renewal | Automatic monthly | Auto-renews per Order Form (typically 12 months) |
| Termination notice | Cancel anytime via the billing portal | 60 days written notice required |
| Signed paperwork required | No (web purchase) | Yes — Order Form, or Purchase Order accepted in lieu |
| Payment method | Credit card via Stripe | Wire transfer, ACH, or invoice |
| Payment timing | Charged upfront, before service starts | Net 30 — invoice issued, due within 30 days |
| Refund policy | Non-refundable | Non-refundable |
| Billing dispute window | Not specified | 60 days from first statement (per Order Form) |
| Purchase order (PO) acceptance | Not applicable | Accepted in lieu of Order Form |
| How to sign up | Self-serve at [genspark.ai/team_pricing](https://www.genspark.ai/team_pricing) | Contact `enterprise@genspark.ai` |
### 2. Features & Usage {#fc-features}
| Dimension | Team Plan | Enterprise Plan |
|---|---|---|
| Zero-credit chat models (top-tier) | Unlimited until Dec 31, 2026 (latest SOTA lineup on the [pricing page](https://www.genspark.ai/team_pricing)) | Included; standard credit pricing applies |
| Zero-credit image models (top-tier) | Unlimited until Dec 31, 2026 (latest SOTA lineup on the [pricing page](https://www.genspark.ai/team_pricing)) | Included; standard credit pricing applies |
| Video generation models | All current models | All current models |
| Audio generation models | All current models | All current models |
| Storage per seat | 60 GB AI Drive (fixed) | Custom capacity (per Order Form) |
| Commercial use rights for AI content | Granted, valid through Dec 31, 2026 | Granted; confirmed in Order Form |
| Credit rollover | Credits do not roll over month to month | Credits do not roll over month to month |
| Admin visibility into seats | Per-seat credit usage and account balance | Same, plus exportable credit usage logs and login history |
| Admin access to member content | None — admins cannot view individual project history or content | None — same posture |
| Additional / multiple subscriptions | Allowed (you can purchase additional team subscriptions or seat blocks) | Allowed via separate Order Forms |
### 3. Admin Controls & Governance {#fc-admin}
| Dimension | Team Plan | Enterprise Plan |
|---|---|---|
| Admin scope | Centralized admin: billing, seats, member roles, usage analytics, connectors | Same baseline; plus custom governance per Order Form |
| SSO / SAML providers | Microsoft Entra ID (Azure AD), Google Workspace, Okta, Generic SAML 2.0 | Same; additional provider integrations available on request |
| Multi-factor authentication (MFA) | Enforced via your IdP | Enforced via your IdP |
| Role-based access control | Standard Admin / Member roles | Agent-level permissions |
| Connector management | Org-wide enable / disable | Org-wide enable / disable |
| Invoicing | Stripe-issued invoices, downloadable from dashboard | Standard accounts payable invoicing (Net 30) |
| Agent governance | None | Enable / disable agents organization-wide |
| Storage quotas | Fixed 60 GB per seat | Custom (per Order Form) |
### 4. Compliance & Certifications {#fc-compliance}
| Dimension | Team Plan | Enterprise Plan |
|---|---|---|
| SOC 2 Type II | Certified | Certified |
| ISO 27001:2022 | Certified | Certified |
| GDPR compliance | Program in progress | Program in progress |
| Trust Portal access | Available — [app.vanta.com/mainfunc.ai/trust](https://app.vanta.com/mainfunc.ai/trust/9mmidu6lo2p5p3z7h1etc){rel="nofollow noopener noreferrer"} | Available, plus detailed audit reports under NDA |
| Penetration test reports | Available via Trust Portal | Trust Portal reports + customer-specific reports on request |
| Custom compliance addendums | Not available | Available (e.g., ITAR/EAR, BAA, industry-specific) |
| Customer audit rights | Annually, with 30 days notice; cost capped at 5% of fees (per agreement) | Annually, with 30 days notice; cost capped at 5% of fees (per Order Form) |
### 5. Data Governance & Privacy {#fc-data-privacy}
| Dimension | Team Plan | Enterprise Plan |
|---|---|---|
| Use of customer data for AI training | Never used to train AI models | Never used to train AI models |
| Subprocessor training guarantees | OpenAI, Anthropic, Google bound by no-training agreements via DPA | Same |
| Platform service improvement | Disabled by default; opt-out available | Disabled by default; opt-out available |
| Aggregated usage analytics | Only if you explicitly enable; anonymized patterns only | Only if you explicitly enable; anonymized patterns only |
| Raw content access for analytics | Never accessed or analyzed | Never accessed or analyzed |
| What Genspark collects | Metadata only — token counts, request counts, performance metrics, auth events | Metadata only — same scope |
| Customer data ownership | Customer retains all rights to customer data | Customer retains all rights to customer data |
| Genspark IP rights | Genspark retains all service / software IP | Genspark retains all service / software IP |
| Data Processing Agreement (DPA) | Not currently available | Available. Custom DPA tailored to compliance needs |
| Submission of personal data | Requires DPA before submission | Requires DPA before submission |
### 6. Data Retention & Deletion {#fc-retention}
| Dimension | Team Plan | Enterprise Plan |
|---|---|---|
| Account data after termination | Deleted 30 days after termination | Deleted 30 days after termination (extension available on request) |
| On-demand data deletion | Available | Available, with certified deletion on request |
| Backup retention after deletion | 90 days | 90 days |
| Legal hold capability | Not available | Available, for litigation support |
### 7. Security Architecture {#fc-security}
| Dimension | Team Plan | Enterprise Plan |
|---|---|---|
| Encryption at rest | AES-256-GCM | AES-256-GCM |
| Encryption in transit | TLS 1.3 preferred, TLS 1.2+ supported | TLS 1.3 preferred, TLS 1.2+ supported |
| Network isolation | Shared infrastructure | Dedicated VPC option (on request, per Order Form) |
| Data residency | US (US-West) by default | Configurable: US, EU, or APAC |
| Security incident response | Standard customer notification | 24-hour notification (per Order Form) + dedicated security contact |
| Vulnerability management | Continuous scanning | Continuous scanning + customer notification of findings |
### 8. Support & SLA {#fc-support}
| Dimension | Team Plan | Enterprise Plan |
|---|---|---|
| Support channel | Email | Dedicated Customer Success Manager + priority queue |
| Critical-issue response time | 24 hours | 4 hours (per Order Form) |
| High-priority response time | 48 hours | 8 hours (per Order Form) |
| Support hours | Business hours (Mon–Fri) | 24/7 for critical issues |
| Uptime SLA | Not specified in Team MSA | 99.9% (per Order Form) |
| Service credits for downtime | Not available | Available per Order Form |
| Dedicated account manager | No | Yes |
| Quarterly business reviews | No | Yes |
| Onboarding | Self-service documentation | Dedicated implementation team |
| Vendor onboarding paperwork | Not provided | Provided on request |
## FAQ {#faq data-section-type=faq}
### How do I choose between Team and Enterprise? {#faq-choose}
If your team is **2–150 people** and you're comfortable with monthly self-serve credit-card billing, **Team** is the right starting point. If you have **151+ users**, need a **custom contract**, **invoice-based billing**, **dedicated support**, **custom data residency / DPA**, or **advanced governance** (agent-level permissions, billing-only roles), choose **Enterprise**. For Enterprise, contact `enterprise@genspark.ai`.
### Do team members need to be online at the same time to use Genspark? {#faq-concurrency}
No. Each team member has a fully independent Genspark account. They can sign in, run agents, save projects, and use their own credits and storage without depending on any other member being online. See [Member Experience](#member-experience) for details.
### Are credits shared across the team? {#faq-shared-credits}
By default, **no** — each seat receives its own monthly credit allocation (12,000 on Team, 25,000 on Enterprise) for that individual member. Credits cannot be transferred between members. To give a specific member extra credits, see [Power user credits](#faq-power-user) below. *(Enterprise customers can negotiate alternative arrangements like shared credit pools in their Order Form.)*
### One of our power users may exceed their monthly credits — can teammates' surplus be reassigned? {#faq-power-user}
By default, **no** — surplus credits do not flow between members. The supported workaround is **Credits Packs**: an admin can purchase packs from **Billing → Add Credits** and either let members self-claim them when they run out, or pre-assign packs to specific power users from **User analytics → Member Stats** by clicking **Assign credit pack**. Enterprise customers can also negotiate a shared credit pool in their Order Form.
### What happens to a member's unused credits when they're removed? {#faq-removed-credits}
Unused credits are **not transferred** to the admin or to other members on either plan. To avoid losing the seat cost, you can invite a new member onto the same seat — they will receive a fresh monthly allocation on their account.
### Can I add members in the middle of a billing cycle? {#faq-mid-cycle}
Yes on both plans. On Team, when you increase your seat count, the additional seats take effect immediately and you're charged a **prorated** amount based on the time remaining in your current billing cycle. Each new member receives their full monthly credit allocation right away. On Enterprise, mid-term seat changes are coordinated through your CSM and reflected in the next invoice or as a true-up per your Order Form.
### Is the 60 GB AI Drive storage shared across the team or allocated per seat? {#faq-storage}
On **Team Plan**, each member has a dedicated **60 GB** of AI Drive storage — it is not pooled or shared with other members. On **Enterprise**, storage capacity is custom and configured in your Order Form. Admins cannot view or access another member's AI Drive contents.
### Are there any fair-use policies on the unlimited zero-credit chat and image models? {#faq-fair-use}
Yes. The "unlimited" promotion on top-tier chat and image models on Team Plan means there are no monthly quotas or credit caps for normal use. To keep the service fast and fair for everyone, we apply guardrails against:
- Automatically or programmatically extracting data in bulk
- Sharing your account credentials or making your account available to others
- Reselling access or using Genspark to power third-party services
We also apply session-based rate limits within a defined time window. Most users won't hit them in normal use; once the window resets, full usage resumes.
### Can I have more than one admin? {#faq-multi-admin}
Yes. Both plans support multiple admins. Promote any member to admin from the **Members** tab.
### Can a billing admin role exist without consuming a seat? {#faq-billing-admin}
Not on Team — every admin counts as a paid seat. **Enterprise** supports billing-only roles configured per your Order Form.
### How do team members log in — with passwords or SSO? {#faq-login}
Both methods are supported, and they can be used together or exclusively per your IdP configuration:
- **SSO** — If your organization configures SAML SSO under [SSO setup](#sso-setup), members in your domain authenticate through your IdP (Microsoft Entra ID, Google Workspace, Okta, or Generic SAML 2.0). You can optionally **require SSO for all members**, which blocks every other login method
- **Email + password / Google sign-in** — Without SSO, each member manages their own login credentials on their personal Genspark account. Genspark does not provision or store passwords centrally on behalf of the admin
### Can a member also keep a personal Plus or Pro subscription? {#faq-personal-sub}
We recommend canceling any personal Plus or Pro subscription **before joining the team**. If the same email keeps both subscriptions active, you'll be billed for both, and each plan's monthly credit allocation will be issued as a separate credit grant on the same account — the two subscriptions don't merge or discount each other, but the credits do pool into a single spendable balance on the account.
### What can admins see about individual members? {#faq-admin-visibility}
Admins **can** see, on a per-member basis:
- Current credit balance and credit-pack claims
- Credit usage breakdown for the last 7 / 30 days or all time, exportable as CSV
- Account activity such as a seat being added or removed
- *(Enterprise only)* Login history and exportable credit usage logs
Admins **cannot** see member project history, files, AI Drive contents, prompts, or any AI-generated outputs. Members' work is private to them.
### Can I disable a connector for some members but not others? {#faq-connector-scope}
Not at this time. Connector settings apply to the whole organization on both plans — if an admin disables a connector, it will not be accessible for any member.
### What payment methods are accepted? {#faq-payment-methods}
- **Team Plan** — Credit card only, processed via Stripe.
- **Enterprise Plan** — Wire transfer, ACH, or invoice; Net 30 terms. Purchase orders are accepted in lieu of an Order Form.
### Can I be billed in a currency other than USD? {#faq-currency}
No — both Team and Enterprise are billed in **USD** by default. The displayed price reflects the base subscription rate; your final charge may include local sales tax (VAT/GST) collected by tax authorities and currency conversion or foreign-transaction fees applied by your bank — these are not charged by Genspark. If your organization has a specific currency requirement, contact `enterprise@genspark.ai` to discuss.
### Can we reference a Purchase Order (PO) number on our invoices? {#faq-po-numbers}
PO numbers are accepted on **Enterprise** invoices and can be used in lieu of an Order Form, subject to standard terms. PO referencing is not supported on Team Plan, which is paid by credit card via Stripe.
### Can a single payment source cover the entire team (e.g., one company card for 30–50 people)? {#faq-single-payment}
Yes — that is the default model. The Admin's payment method on file (credit card on Team, wire / ACH / invoice on Enterprise) covers all seats in the team subscription. Members do not need their own payment method to use the team.
### Can Genspark provide written confirmation of plan terms for our procurement / IT team? {#faq-procurement-docs}
Yes. For procurement-ready documentation — including independent member access, credit allocation, storage terms, and security posture — contact `enterprise@genspark.ai`. They can also coordinate any custom paperwork needed for vendor onboarding.
### Are my prompts and content used to train AI models? {#faq-training}
No. Both Team and Enterprise accounts are automatically opted out of model training. Per Genspark's Data Processing Agreement (DPA), your prompts and content are never used to train Genspark models. Our subprocessors (OpenAI, Anthropic, Google) are bound by no-training agreements, with **Zero Data Retention (ZDR)** in place where applicable.
### Is a custom Data Processing Agreement (DPA) available? {#faq-dpa}
- **Team Plan** — DPA is not currently available
- **Enterprise Plan** — A custom DPA tailored to your compliance requirements is available, and can include industry-specific addendums (e.g., ITAR/EAR, BAA)
For DPA requests, contact `enterprise@genspark.ai`.
### How do I get a SOC 2 Type II report? {#faq-soc2}
A SOC 2 Type II report is available to **Enterprise** customers under NDA. Contact your CSM or `enterprise@genspark.ai`.
### Can data be hosted outside the US? {#faq-residency}
Team Plan data is hosted in the US (US-West) by default. **Enterprise** offers configurable data residency in the US, EU, or APAC, defined in your Order Form.
### Are AI-generated outputs allowed for commercial use? {#faq-commercial-use}
Yes. All Team and Enterprise members are granted commercial use rights for AI-generated content — including images, edited images, videos, audio, and other outputs — for client work, deliverables, websites, and advertising creatives. Refer to [genspark.ai/terms](https://www.genspark.ai/terms) for the official policy.
### Does commercial use cover image editing, background removal, and transparent PNG export? {#faq-commercial-editing}
Yes. The commercial-use rights granted to Team and Enterprise members cover all AI-generated content **and** edits derived from it — including image editing, background removal, transparent PNG export, video edits, and audio outputs. Refer to [genspark.ai/terms](https://www.genspark.ai/terms) for the full policy.
### Are there limits on how I can use the AI-generated content commercially? {#faq-commercial-limits}
There are no specific limitations on duration, media type, use case, or distribution beyond Genspark's standard [terms of service](https://www.genspark.ai/terms). Outputs can be used for client work, deliverables, websites, advertising creatives, social media, and other commercial channels.
### I can't access the admin dashboard — what's wrong? {#faq-admin-access}
The management page is only available to members with the **Admin** role. If you believe you should have admin access, ask an existing admin on your team to promote your account. If your team has no remaining admins, contact our support team for assistance — for Enterprise, reach out to your CSM directly.
# How Credits Work {#how-credits-work}
Source: https://www.genspark.ai/helpcenter/credits-guide
## What are credits? {#what-are-credits data-section-type=overview}
This guide explains what credits are, what drives their cost, how to pick the right model for a task, and how to keep your spend under control.
Credits measure computational effort. Every time you generate something — text, an image, a video, or audio — you use credits based on the processing power, memory, and model time the task requires.
## What Affects How Many Credits You Use {#what-affects-how-many-credits-you-use data-section-type=general}
| Factor | Lower cost | Higher cost |
|---|---|---|
| Model tier | Standard tiers | Ultra tiers |
| Text length | Short prompts, short replies | Long context, multi-turn conversations |
| Image resolution | Standard resolution | High-res / 4K |
| Video duration | 4 seconds | 8–16+ seconds |
| Audio length | Short clip | Long narration |
| Number of outputs | 1 version | Multi-model fan-out (MoA) generates N versions |
| Task complexity | Simple Q&A | Research + writing + media in one task |
### What Compounds Your Costs {#what-compounds-your-costs}
Regenerations — each retry costs the same as the original. Refine your prompt or make small manual edits instead of regenerating.
Agent retries — each failed retry costs credits. Attach required files and give clear, well-scoped instructions before you start. If a task keeps failing on the same step, stop it and fix the input.
Long conversations — every message re-processes the full context window. Use in-editor manual tools for small tweaks instead of re-prompting.
Editing vs. creating — small edits cost far less than generating from scratch.
## How to Choose the Right Model {#how-to-choose-the-right-model data-section-type=general}
You don't need to memorize model names to spend wisely.
Use a standard option when:
- You want a reliable everyday default.
- The task needs decent quality but isn't high-stakes.
- Examples: drafting an internal update, creating a presentation outline.
Use a higher-end option when:
- The output is user-facing or high-visibility.
- You're working on a final version rather than a draft.
- Examples: writing polished external copy, complex analysis or coding, generating a final campaign image.
### Practical Examples {#practical-examples}
| Task | Good starting choice | When to upgrade |
|---|---|---|
| Summarize a document or email thread | Standard | Only if the summary misses nuance |
| Draft a blog outline or internal memo | Standard | If the writing needs stronger structure or polish |
| Complex reasoning, coding, or strategy work | Higher-end | Start high if accuracy matters more than speed |
| Generate a poster or polished creative asset | Standard image model | Upgrade for final-quality visuals |
| Create a short test video | Standard video model | Upgrade for realism or final delivery |
| Generate audio or music | Single model first | Use Auto-select only if you want multiple versions |
## How Model Selection Works {#how-model-selection-works data-section-type=general}
Genspark uses two model-selection systems, depending on the product.
Most products use a simple Standard / Ultra quality tier — you choose in the input box before you generate. Standard is the everyday default; Ultra is for final, high-stakes outputs.
Note: these are model-quality tiers — separate from the Cloud Computer subscription plans.
### Products That Let You Pick the Exact Model by Name {#products-that-let-you-pick-the-exact-model-by-name}
A smaller set of products expose the full model list so you can choose exactly what you want — including lower-cost, standard, and high-end options by name: AI Chat, Genspark Code, Claw, AI Image, AI Video, and AgentBase.
| Product | Model selector location |
|---|---|
| AI Chat, Genspark Code, Claw, AgentBase | Button at the bottom of the toolbar |
| AI Image, AI Video | Parameter bar above the prompt input |
AI Audio and AI Music are the exception — they give you both a Standard/Ultra tier selector and a model-by-name picker (Auto-select), so you can choose whichever way you prefer to work.
⚠️ Auto-select multiplies cost: AI Audio and AI Music default to Auto-select, which generates several versions at once — you pay for each. Pick a single model manually to control spend.
### Model Cost Reference {#model-cost-reference}
Model lineups and prices change frequently — check the live pricing inside Genspark. When in doubt, start with a lower-cost option and upgrade only if the output falls short.
## Simple Ways to Use Fewer Credits {#simple-ways-to-use-fewer-credits data-section-type=general}
- Start with a lower-cost option first. Switch to a higher-end model only for the final pass.
- Keep prompts focused. A clear prompt usually reduces unnecessary retries.
- Generate one version before generating many. Confirm the direction first, then ask for variants.
- Avoid multi-model fan-out unless you need it. In Audio and Music, Auto-select multiplies cost.
- Save top-tier models for high-value tasks. Use them when better quality will actually change the outcome.
## Your Credit Balance {#your-credit-balance data-section-type=general}
Open the Credits item in the bottom-left menu to open the /credit-usage page — your current balance, remaining allocation, and refresh date.
The Settings dialog has a Billing section for plan and payment management only; balance and history live on the /credit-usage page.
The usage log can take a few hours to catch up — a task you just ran may not appear right away even though your balance already reflects it.
## When Credits Replenish {#when-credits-replenish data-section-type=general}
| Plan | Refresh schedule |
|---|---|
| Free | 100 credits/day, Free accounts have a lifetime allowance of free credits. Once your account has used its lifetime allowance, the daily free refill stops — your current balance stays and keeps working until it runs out. |
| Plus / Pro | Monthly, at the start of your billing cycle. |
⚠️ Credits don't carry over: Credits reset each billing period. Separately purchased top-up credits are valid for 3 months and carry across cycles.
## Changing Plans and Your Credits {#changing-plans-and-your-credits data-section-type=general}
**Upgrading** — takes effect immediately; you're charged a prorated amount; supplemental credits (the difference) are added right away.
**Downgrading** — current plan stays active until end of billing period; credits don't carry over to the lower plan.
⚠️ Unused credits from a higher plan don't carry to a lower plan. Spend them before your renewal date.
**Canceling** — plan stays active until end of period; account reverts to Free (100 daily credits until lifetime cap is reached) after that; credits from the canceled period aren't refunded.
**After upgrading** — Plus and Pro subscriptions include AI Chat and AI Image core models at no credit cost within the limit. Once this limit is reached you will use your monthly credits. Flagship models use your credits and don't pull from your Free Quota. If you subscribed on or before September 18, 2026 and keep your subscription continuously active, your current benefits continue unchanged through December 31, 2026. Your paid plan's monthly allocation replaces the Free plan's 100 daily credits, you do not get both.
**Free Quota** – Plus and Pro subscriptions include Free quota, a free weekly credit balance to be used on agents with Standard mode (Super Agent, AI Slides, AI Sheets, AI Docs, Deep Research, and more), at no credit cost. It resets every 7 days. When it runs out, tasks switch to credits automatically. See [Membership Plans](https://www.genspark.ai/helpcenter/membership-plans#:~:text=zero%20credit%20cost.-,How%20Free%20Quota%20Works,-Free%20Quota%20is) for more information.
### Changing Your Billing Interval (Monthly ↔ Yearly) {#changing-your-billing-interval}
Yearly plans grant your credit allocation each month across the year, not the full year up front. So "unused credits" on an annual plan always means the current month's credits, just like a monthly plan.
**Monthly → Yearly**: takes effect immediately with a prorated charge. Your credit allocation refreshes right away, and continues to refresh each month across the year.
**Yearly → Monthly**: what happens depends on whether the new monthly plan gives you more or fewer credits than your current plan.
- Same or lower allocation → treated as a downgrade. Your current plan runs to the end of the billing period; the new monthly plan begins at your next renewal.
- Higher allocation → treated as an upgrade. Takes effect immediately with a prorated charge; your new (higher) monthly allocation is issued right away.
## FAQ {#faq data-section-type=faq}
### Do failed generations cost credits? {#do-failed-generations-cost-credits}
It depends on how the generation failed. If your prompt was blocked by the safety filter before anything was generated, your credits are refunded automatically. If a generation completed and the output was then flagged, it's still charged — the model did the work — even though it displays as "failed."
### What doesn't cost credits? {#what-doesnt-cost-credits}
- Browsing, organizing, or viewing files in AI Drive.
- Editing notes or switching views in Dashboards & CRM.
- Generations that are automatically refunded (prompt-blocked failures; not output-flagged NSFW content).
### Why did one task cost more than another? {#why-did-one-task-cost-more-than-another}
Cost scales with model tier, output length/resolution/duration, number of versions, and task complexity.
### Can I get credits back for a failed or low-quality generation? {#can-i-get-credits-back}
Prompt-blocked failures are refunded automatically. Output-flagged NSFW content is still charged. Quality disputes don't get automatic refunds. For technical billing issues, contact support with a link to the task, what went wrong, and your approximate balance before and after.
### Is there a limit to free daily credits? {#Free-flan-credit-limit}
Yes. Free accounts have a lifetime allowance of free credits. Once you've used it, the daily refill stops — your remaining balance keeps working until it runs out. To keep generating, upgrade to a paid plan.
### What happens if I hit the Auto-select limit unexpectedly? {#auto-select-limit}
Auto-select fans out to multiple models and charges for each. Switch to a single model in AI Audio or AI Music to control spend.
### Why did I hit a usage limit when I still have credits? {#usage-limit-with-credits}
AI Chat and AI Image have a limit separate from your credit balance. Very heavy use can trigger a temporary limit based on model spend over a fixed 5-hour window. Once the window starts, it clears entirely when the 5 hours are up — you'll see the reset time when you hit it. Buying more credits won't lift it, and hitting it doesn't cost you any credits. Higher plans have higher limits.
### I subscribed before September 18, 2026. Does anything change for me? {#subscribe-before-September-18}
If your Plus or Pro subscription started on or before September 18, 2026 and stays active without interruption, your current unlimited AI Chat and AI Image benefits continue unchanged through December 31, 2026"
# Troubleshooting Guide
Source: https://www.genspark.ai/helpcenter/troubleshooting-guide
## FAQ {#faq data-section-type=faq}
### What does the "NSFW" message mean? {#nsfw-message}
If you see an "NSFW" notice, it means your generated content didn't comply with the model's content or safety guidelines. The model automatically filters or blocks output that violates these policies—this isn't a system error.
This is by design to ensure safe, responsible AI use.
### Why do Sora video generations sometimes fail? {#sora-generation-fail}
Due to occasional OpenAI API instability, Sora model generations may fail from time to time.
**Higher failure rates occur with:**
- Content involving human faces
- Live-action video content requests
**Remember:** No credits are deducted for failed generations, so feel free to try again!
### Why are some AI models removed from Genspark? {#models-removed}
We regularly update our model lineup to offer the most advanced AI technology available. Older models are retired when they can no longer be maintained to our quality standards or when superior alternatives exist. This is standard practice across leading AI platforms and ensures you always have access to the best-performing models.
### Where do the open-source models on Genspark run? {#open-model-hosting}
Several of the models Genspark offers are open-source, including DeepSeek, Kimi, GLM, MiniMax, and Qwen. Open-source means the model weights are published, so anyone can run the model on their own infrastructure — and that is what we do. Genspark runs these models on inference platforms based in the United States, including **Fireworks AI** and **Baseten**.
This means the company that originally built the model never sees your requests. Your prompts, files, and results are sent to the US platform hosting the model, not to the model's developer.
We also hold Zero Data Retention (ZDR) agreements with these providers: they do not keep your prompts or results once your request has been answered, and nothing you send is ever used to train a model.
### What does a 401 Error mean? {#error-401}
You're seeing this because your login session has expired or there's an authentication issue. Don't worry—it's just our way of keeping your account secure.
**How to fix it:**
1. Log out and log back in — This usually solves the problem
2. Clear your browser cache and cookies for [Genspark.ai](https://www.genspark.ai/)
3. Try an incognito/private browsing window
4. Verify your login credentials — Make sure you're using the correct username and password
5. Check if your account is active — Ensure it hasn't been suspended
6. Disable browser extensions — Some extensions can interfere with authentication
7. Reset your password — If you're unsure about your credentials
**Still can't get in?**
[Submit a ticket](#support){.helpcenter-support-link} with your username (never include your password). We'll help you get back in quickly.
### What does a 400 Error mean? {#error-400}
This error happens when there's a formatting issue, missing information, or incompatibility between your request and our system. Don't worry—this isn't something you did wrong.
**You might see this when:**
- Form submissions fail
- You get "Invalid request" messages
- Pages won't load after clicking submit
- File uploads don't go through
- Search functions return errors
**How to fix it:**
1. Refresh the page and try again
2. Clear your browser cache and cookies
3. Try a different browser
4. Sign in with a different account to see if the issue persists
**Still stuck?**
Tell us exactly what you were trying to do, and we'll help you sort it out.
### What does a 500 Error mean? {#error-500}
This means we're experiencing a temporary server issue. This is on us, not you—nothing you did caused this.
**How to fix it:**
1. Wait a few minutes and refresh the page
2. Clear your browser cache
3. Try again later
4. Log in with a different account to see if the issue persists
Our technical team gets automatically notified when this happens and is already working on it. Thanks for your patience!
### What does a 1015 Error mean? {#error-1015}
Our security system detected unusually high activity from your connection and temporarily limited access to protect our servers. This is temporary and will resolve itself.
**How to fix it:**
1. Wait 10-15 minutes — The limit will lift automatically
2. Refresh the page after waiting
3. Clear your browser cache and cookies
4. Restart your internet connection or try a different network
5. Disable VPN/proxy if you're using one
**Why this happens:**
- Too many requests sent in a short time
- Shared network with heavy usage (office, school, coffee shop)
- Browser extensions causing rapid requests
- Network connectivity issues
**Still blocked after 30 minutes?**
Contact our support team—we can help identify what triggered the protection or whitelist your IP. Include the time this happened and your approximate location.
Your account is safe. This is just our way of keeping the service running smoothly for everyone
### How can I help troubleshoot my issue? {#troubleshoot-help}
When you report an issue, sharing these details helps our engineering team identify and fix problems faster:
**Information that helps:**
- Webpage link (if on desktop)
- Session link with the issue, screenshot, or screen recording
- Login email (never your password)
- Device & OS version (Windows, Mac, iPhone, iPad, iOS, Android, etc.)
- Browser or app version
- Browser extensions (if applicable) — Screenshot from chrome://extensions/ or edge://extensions/
- Specific feature you were using when the issue occurred
**How to share a session link:**
1. Click the task title to open the project
2. Click the sharing icon in the upper right corner
3. Invite support@genspark.ai as a **Viewer**
4. Copy the project link and include it in your message to support
**Your privacy matters:**
We understand your concern about personal data. Only share what you're comfortable with—we just want to help identify and solve the problem.
### I lost my project, where did it go? {#lost-project}
1. Open **My Library** from your profile/avatar menu.
2. **It opens on the Bookmark tab — click over to the "History" tab.**
3. Your projects are listed **newest first, grouped by date**.
4. **Keep scrolling** — older projects load as you scroll down. The first screen only shows your most recent work.
**Tip:** If you just created it, **give it a moment and refresh** — the title can take a few seconds to appear.
# Connectors & Integrations {#connectors-integrations}
Source: https://www.genspark.ai/helpcenter/connectors-and-integrations
## What are Connectors & Integrations? {#what-are-connectors-and-integrations data-section-type=overview}
Connectors let Genspark work directly with the apps and services you already use, so it can read information and perform actions without you copying data back and forth. Connect a service from the Connectors tab in Skills, and Genspark products with connector support can use that connection, subject to your permissions and organization settings.
## Quick Start {#quick-start data-section-type=overview}
1. **Open the Connectors tab.** In Skills, navigate to Connectors.
2. **Find and connect a service.** Search or browse for the service and choose the intended connection method. For OAuth, sign in to the provider and review the consent screen. Arrange administrator consent if your tenant requires it. For an API Key, token, or MCP URL, prepare the required credential with the provider and enter it in the setup form.
3. **Check the connection.** Return to the Connectors tab and confirm that the service shows as Connected. For a setup flow that offers a connection test, complete that test as well.
4. **Use it.** Start with a read task on a resource you know you can access. If your workflow needs sending, creating, or updating, verify those permissions separately.
```trythisprompt
Summarize unread messages in my connected work Gmail account.
```
```trythisprompt
Find Salesforce opportunities expected to close this month.
```
```trythisprompt
Summarize last month's Ramp spending by department.
```
## Connector Reference {#connector-reference data-section-type=general}
The reference below covers 37 standard connection options, including the Google Suite and Microsoft 365 bundles. Additional connections that require allowlist access or a customer pilot are listed separately.
Availability depends on your account, organization settings, and the provider's requirements. The tasks listed are examples; available actions depend on the connection's tools and granted permissions.
**Connection type** describes how Genspark calls a service:
- **API Call:** a native integration uses the provider's API.
- **MCP:** Genspark uses tools exposed by an MCP server.
**Authentication** describes how you grant access. MCP connections can use OAuth, an API Key, an access token, or a dedicated server URL.
In the scope-options column:
- **Fixed request** means the connection requests a predefined permission set. Genspark does not provide individual scope checkboxes.
- **Read-only policy** means the connection supports narrowing requested permissions when your organization's read-only feature is enabled. See [Permissions and Access](#permissions-access) for the limits.
- **Provider configuration** means you select resources or configure permissions in the provider's app, account, token, or server.
### Communication {#communication}
| Service | Connection type | Authentication | Scope options | What you can do |
| --- | --- | --- | --- | --- |
| Gmail | API Call | OAuth 2.0 | Fixed request; read-only policy | Read, search, draft, send, and manage email. The current grant also includes contact reading. |
| Outlook Email | API Call | OAuth 2.0, delegated access | Fixed request; read-only policy | Read, search, send, and manage Outlook email, including authorized shared-mailbox access. |
| Slack | API Call | OAuth 2.0, user authorization | Fixed request; read-only policy | Search conversations, read channels, and send messages. |
| Microsoft Teams | API Call | OAuth 2.0, delegated access | Fixed request; read-only policy | Read chats and channels, send messages, and access team information within your permissions. |
| Google Chat | API Call | OAuth 2.0 | Fixed request; read-only policy | Read spaces and messages, and send replies. Connect Google Chat separately from Google Suite. |
| LINE Send Message | API Call | Account linking through LINE Login | Fixed identity permissions | Receive messages from the official bot on your linked LINE account. This does not authorize reading personal chat history. |
| X (Twitter) | API Call | OAuth 2.0 with PKCE | Fixed request; read-only policy | Search and publish posts, and work with permitted social actions. |
| LinkedIn Lite | API Call | Manually supplied OAuth access token | Provider configuration | Read your profile and draft or publish posts. Beta; the supplied token does not refresh automatically. |
### Productivity {#productivity}
| Service | Connection type | Authentication | Scope options | What you can do |
| --- | --- | --- | --- | --- |
| Google Suite | API Call | OAuth 2.0 | Connect services together or individually; read-only policy | Connect Gmail, Calendar, and Drive together. Docs, Sheets, and Slides file access is provided through Drive. |
| Google Calendar | API Call | OAuth 2.0 | Fixed request; read-only policy | View schedules and create or manage calendar events. |
| Google Drive | API Call | OAuth 2.0 | Fixed request; read-only policy; file authorization | Search and read files, and work with Google Docs, Sheets, and Slides files. Write access applies to files created by the app or authorized by you. |
| Google Contacts | API Call | OAuth 2.0 | Fixed read-only request | Search and view contact details. |
| Microsoft 365 | API Call | OAuth 2.0, delegated access | Connect services together or individually; read-only policy | Connect mail, calendars, files, Teams, profile, and contacts. Outlook contact access is included in this bundle. |
| Outlook Calendar | API Call | OAuth 2.0, delegated access | Fixed request; read-only policy | Find, create, and manage calendar events. |
| Microsoft OneDrive | API Call | OAuth 2.0, delegated access | Fixed request; read-only policy | Search, read, and analyze files and folders. |
| Microsoft SharePoint | API Call | OAuth 2.0, delegated access | Shared file permission set with OneDrive; read-only policy | Search sites and access pages, lists, and files. This connection does not offer per-site scope selection. |
| Box | API Call | OAuth 2.0 | App-defined permissions; read-only policy | Find files and folders, create folders, and access file information. |
| Confluence | API Call | Atlassian OAuth 2.0 | Fixed request shared with Jira; read-only policy | Search and read pages and spaces, and prepare page updates. |
| Notion | API Call | OAuth 2.0 | Provider configuration: authorize pages | Read, create, and edit pages and content authorized for the Notion integration. |
| Notion MCP | MCP | OAuth 2.1 with PKCE | Provider authorization; no Genspark scope picker | Use Notion's hosted tools within the connecting user's Notion permissions. This is separate from the native Notion connection. |
| ClickUp | MCP | OAuth 2.1 with PKCE | Provider authorization; no Genspark scope picker | Find tasks, review workload, create tasks, and update statuses. |
| monday.com | MCP | OAuth 2.1 with PKCE | Provider authorization; no Genspark scope picker | Read boards and items, create items, and update statuses. |
### CRM & Sales {#crm-sales}
| Service | Connection type | Authentication | Scope options | What you can do |
| --- | --- | --- | --- | --- |
| Salesforce | API Call | OAuth 2.0 | Fixed request; configure roles and object permissions in Salesforce | Read and update authorized CRM records, contacts, and opportunities; analyze the sales pipeline. |
| HubSpot | API Call | Private App Token | Provider configuration: object-level read/write scopes | Work with CRM contacts, companies, and deals. Create the token in HubSpot before connecting. |
| Pipedrive | API Call | OAuth 2.0 | Provider app configuration; no Genspark scope picker | Search and manage deals, contacts, organizations, and activities. |
| Mailchimp | API Call | OAuth 2.0 | Provider app and account permissions; no Genspark scope picker | Review audiences, campaigns, subscribers, and campaign results. |
| Affinity CRM | MCP | API Key sent as a Bearer token | Provider configuration; no OAuth scope picker | Explore relationship intelligence and CRM data through Affinity's tools. |
### Developer {#developer}
| Service | Connection type | Authentication | Scope options | What you can do |
| --- | --- | --- | --- | --- |
| GitHub | API Call | OAuth App authorization | Fixed request; read-only policy narrows only some scopes | Search repositories and issues, and work with repository content. The repository grant still includes write access. |
| Jira | API Call | Atlassian OAuth 2.0 | Fixed request shared with Confluence; read-only policy | Search, create, and update Jira work items. |
| Linear | MCP | OAuth 2.1 with PKCE | Provider authorization; no Genspark scope picker | Find, create, and update issues, projects, and comments. |
### Data {#data}
| Service | Connection type | Authentication | Scope options | What you can do |
| --- | --- | --- | --- | --- |
| Airtable | MCP | OAuth 2.1 with PKCE | Provider authorization; no Genspark scope picker | Query and update records, fields, and views in authorized bases. |
### HR & Recruiting {#hr-recruiting}
| Service | Connection type | Authentication | Scope options | What you can do |
| --- | --- | --- | --- | --- |
| Greenhouse | MCP | OAuth 2.1 with PKCE | Provider-defined scope and administrator configuration | Search jobs, candidates, applications, and interview stages. Your administrator may need to enable MCP access. |
| Ashby | MCP | OAuth 2.1 with PKCE | Connecting user's provider permissions | Review candidates, jobs, applications, interviews, and feedback. Your administrator may need to enable MCP access. |
### Operations & Finance {#operations-finance}
| Service | Connection type | Authentication | Scope options | What you can do |
| --- | --- | --- | --- | --- |
| Stripe | MCP | OAuth 2.1 with PKCE | Provider authorization; no Genspark scope picker | Work with payment data, subscriptions, products, and customers, subject to the authorized tools and permissions. |
| PayPal | MCP | OAuth 2.1 with PKCE | Provider authorization; no Genspark scope picker | Work with invoices, orders, payments, and transactions, subject to the authorized tools and permissions. |
| Ramp | MCP | OAuth 2.1 with PKCE | Dedicated read-only scope restriction | Read spend, card, transaction, bill, and reimbursement data. Money-movement write permissions are excluded. |
| Zapier | MCP | Dedicated MCP server URL containing credentials | Configure enabled tools and app connections in Zapier | Run the actions enabled on your Zapier MCP server. Treat the connection URL as a secret. |
### Community {#community}
Custom and community MCP servers are governed separately by your organization's custom MCP policy. Their tools, authentication, and resource access depend on the server operator. They are not additional standard catalog entries in the reference above. Check with your administrator before relying on a server for an enterprise workflow.
### Early Access and Customer Pilots {#early-access}
The following connections require allowlist access or a customer pilot. They are not enabled for every production account. Once enabled, organization policies and provider requirements still apply.
| Service | Connection type | Authentication | Scope options | What you can do |
| --- | --- | --- | --- | --- |
| Salesforce Sandbox | API Call | OAuth 2.0 | Fixed request; provider roles and object permissions | Test CRM workflows in a Salesforce sandbox. This authorization is separate from the production Salesforce connection. |
| Daloopa Financial Data | MCP | Provider-managed sign-in; confirm the flow during onboarding | Confirm provider permissions during the pilot | Retrieve company fundamentals, metrics, and financial statements after provider sign-in and connection validation. |
| Intercom | MCP | OAuth 2.1 with PKCE | Provider authorization; no Genspark scope picker | Find support conversations and contacts, and summarize customer feedback. |
| Amplitude | MCP | OAuth 2.1 with PKCE | Provider authorization; no Genspark scope picker | Explore funnels, retention, events, and user segments. |
| Snowflake | MCP | Programmatic Access Token (PAT) | Configure Snowflake roles and object grants | Query or search data through your Snowflake-managed MCP server. Supply the account, database, schema, MCP server, and PAT. The current connection uses PAT authentication. |
| SimilarWeb | MCP | API Key | Provider account and key permissions | Explore website traffic, keywords, and audience data available to your provider account. |
| HubSpot MCP | MCP | OAuth 2.1 with PKCE | Provider authorization; no Genspark scope picker | Use HubSpot's hosted CRM tools. Install separately from the standard native Private App Token connection. |
## Permissions and Access {#permissions-access data-section-type=general}
### What can I choose during authorization? {#permission-options}
Most OAuth connections request a predefined permission set. You can choose the provider account and, for Google and Microsoft, connect individual services instead of using a bundle.
Native Notion lets you authorize pages. For customer-managed tokens such as HubSpot's Private App Token, configure the needed permissions with the provider before entering the token in Genspark.
MCP OAuth permissions usually follow the server's authorization metadata and consent screen. Genspark does not offer individual scope checkboxes for these connections. A provider's account roles, resource sharing, and available tools can further limit access.
### Important permission boundaries {#permission-boundaries}
- **Google Suite:** the bundle requests Gmail, Calendar, and Drive access. Google Chat is a separate connection. Gmail also requests contact-reading permission and, in read/write mode, Gmail settings permission.
- **Google Drive:** `drive.readonly` permits broader reading, while `drive.file` permits writing files created by the app or authorized by you. Selecting files for write access does not restrict all read access to those files.
- **Microsoft 365:** connections use the signed-in user's delegated permissions. They do not grant an administrator role. OneDrive and SharePoint request the same file permission set; this flow does not offer site-by-site scope selection.
- **Notion:** the native integration's authorized-page boundary and Notion MCP's user-based access are different. Review the option you connect.
- **Jira and Confluence:** both use the same Atlassian authorization and request permissions for both services.
- **GitHub:** the `repo` scope includes repository write access even when the organization read-only policy narrows other scopes.
- **Ramp:** requests are limited to the intersection of 21 allowed read-only scopes and the server's declared scopes. The connection excludes money-movement writes, bank account number access, and spendable card credentials.
### Organization controls and read-only access {#organization-controls}
Enterprise administrators can allow a connector for all members, restrict it to selected teams, or disable it. These settings control who can use a connection. Provider permissions control which data and operations that connection can access.
If the read-only management feature is enabled for your organization, supported OAuth connections can request narrower permissions. Google, Microsoft, Slack, Atlassian, Box, and X support this; GitHub supports only partial narrowing. Native Notion, Salesforce, Pipedrive, Mailchimp, and supplied tokens do not automatically become provider-issued read-only credentials.
Read-only restrictions on Genspark tools and the permissions held by a provider token are separate controls. Turning on the policy does not automatically remove all permissions from previously issued tokens. Your administrator may need to arrange reauthorization, change provider permissions, or rotate credentials. Direct access through a credential-bearing CLI or sandbox requires separate assessment.
MCP connections do not have a blanket read-only guarantee. Check provider scopes, account roles, resource access, and server tools. For Zapier, also review the underlying connected apps. Ramp's dedicated scope restriction applies separately.
## Managing Connections {#managing-connections data-section-type=general}
### Multiple email accounts {#multiple-email-accounts}
Gmail and Outlook Email support multiple connected accounts. Open the connector's details to view the account list, add another account, or disconnect an account using its own row.
The list identifies the default account and can show when reauthorization is needed. Other connectors do not necessarily offer the same account-list controls. If you have several mailboxes connected, specify the intended account in your request.
### Disconnect an account or a service {#disconnecting}
- **One Gmail or Outlook account:** use that account's row to remove its mailbox connection and account-based calendar/contact access. Other mailbox accounts remain connected.
- **Whole Google or Microsoft connection:** the whole-connector disconnect action affects related services that share the authorization, and removes bound mailboxes from AI Inbox/GenMail. Review the affected services before confirming. Microsoft's whole-connection flow also attempts provider session revocation; if successful, other Microsoft apps may require sign-in again.
- **Jira or Confluence:** disconnecting either disconnects both because they share an authorization.
- **MCP:** disconnecting removes the MCP registration from Genspark. Provider authorization or customer-created credentials may need to be revoked separately.
Removing one Outlook account does not revoke Microsoft sessions. Removing one Gmail account attempts provider revocation only when other related connectors no longer reference that identity.
For complete offboarding, review the provider's connected-app settings and revoke or rotate customer-created tokens as needed. Disconnecting does not delete the provider account itself.
## Troubleshooting {#troubleshooting data-section-type=troubleshooting}
### Why can't I find a connector I expected to see? {#missing-connector}
The connector may be in early access or restricted by your organization's settings. Administrators can limit which connectors are available to members or teams.
An already connected service may remain visible as blocked so you can disconnect it. A service appearing in a pilot list does not mean your account has been enabled.
### Why is the connector connected but a file or action unavailable? {#connected-but-unavailable}
Connected means the connection was established; it does not guarantee access to every resource or operation.
Check the selected account, the provider's resource-sharing settings, and the required read or write permissions. For MCP, also check that the server exposes a suitable tool. Ask your administrator to review any organization restriction before reconnecting with different permissions.
### Why am I being asked to reconnect? {#reconnect}
A provider grant may expire, be revoked, or no longer include the required permissions. Reconnect the intended account when prompted.
Customer-managed keys and manually supplied tokens may need replacement; LinkedIn Lite does not currently refresh its supplied OAuth token automatically.
### What should I gather before contacting support? {#contacting-support}
- The connector name and category, and whether you selected API Call or MCP.
- The exact error message, the step that failed, and the time and time zone.
- Whether you used OAuth, an API Key, a token, or a dedicated MCP URL.
- Your Genspark plan tier and any organization restriction shown.
- The relevant provider workspace or tenant, if needed to identify the connection.
Remove access tokens, API Keys, and credential-bearing URLs from screenshots and logs before sharing them with support.
## FAQ {#faq data-section-type=faq}
### Why do I see more than one HubSpot card in the Connectors panel? {#hubspot-connection-methods}
The standard native HubSpot connection uses a Private App Token that you create in HubSpot with the required CRM object permissions.
HubSpot MCP is a separate OAuth connection that requires allowlist access. Choose the connection available to your organization and approved by your administrator.
### Do I need a paid subscription for any connector? {#provider-requirements}
Requirements vary by provider. Some services need an eligible plan, API or MCP access enabled, or administrator consent. Greenhouse and Ashby may require an administrator to enable MCP access.
Check the provider's current requirements during setup; a visible Genspark card does not establish that your provider account is eligible.
### Can LINE read my personal conversations? {#line-messages}
The listed connection is LINE Send Message. It links your identity and lets the official bot send messages to your linked account. It does not grant access to your personal chat history.
### Is every MCP connection read-only? {#mcp-read-only}
No. Operations depend on provider scopes, roles, and exposed tools. Ramp has a dedicated read-only scope restriction.
For other services, check the actual authorization and server configuration, including connected downstream apps for Zapier.
