Announcing Genspark MoA v2: Top Performance with Gemini 2.0!

We're excited to announce the release of Genspark MoA v2 (Mixture-of-Agents) now incorporating the newly released Gemini 2.0 and achieving a remarkable top score in our latest benchmarks!

You can try it now in the Genspark Chat Agent

Below is a comprehensive benchmark comparison chart that showcases these outstanding results. The data clearly demonstrates that Genspark MoA v2 achieves elite-level performance while maintaining remarkable resource efficiency.

deep-seek-v1*: removed response inside to make sure output is comparable; o3-mini**: set reasoning_effort="high"

We are committed to continuously improving our models. Your feedback is invaluable as we strive to provide the best possible experience.