Google introduced Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking for more natural real-time voice interaction, visual context, and complex task execution. Google positions the standard Live variant for scalable, cost-sensitive voice agents and the Extended Thinking variant for higher-complexity multi-step reasoning; both are available through the Gemini API, with broader use across Google Workspace, Search, and the Gemini app. The post also reports third-party and internal benchmark scores, which depend on each benchmark and test setup and should not be treated as universal workload performance.

Editor's note

This summary is based on the primary source. Open the original for full context, data, and qualifying details.