The gist: Google’s new speech model Gemini 3.1 Flash Live improves accuracy and response speed for more natural real-time voice conversations.
Google has released Gemini 3.1 Flash Live, a speech model with improved accuracy and lower latency for more natural real-time conversations. Developers and enterprises can use the model via the Gemini Live API.
Gemini 3.1 Flash Live is Google’s current speech model for high-quality real-time audio interactions. The model was specifically designed to enable reliable and naturally sounding conversations. Key improvements include higher accuracy while simultaneously reducing latency, resulting in smoother and more precise voice interactions.
Developers can access the model via the Gemini Live API, which is available in Google AI Studio. For enterprise customers, Google thereby offers a way to improve customer experiences in the voice interaction domain. The general public can test the model via Search Live and Gemini Live, which is now available in more than 23 countries.
The release is represented by Google’s product manager Valeria Wu and software engineer Yifan Ding from the Gemini group.
Source: ainews-dev.lumi-systems.io · Published May 17, 2026
Lumi AI News — AI-assisted curation pursuant to Art. 50 EU AI Act. Paraphrase and classification by Lumi News Pipeline v1.5.2.