Gemini 3.8 Live Extended Thinking: World’s Best Voice AI
Google releases Gemini 3.8 Live & Gemini 3.8 Live Extended Thinking (voice AI models) which support 97 languages. Gemini 3.8 Live Extended Thinking achieves an 82.6 score on the Artificial Analysis Speech to Speech Quality Index.
Google announced two new voice AI models called Gemini 3.8 Live & Gemini 3.8 Live Extended Thinking. Gemini 3.8 Live Extended Thinking achieved an 82.6 score on the Artificial Analysis Speech to Speech Quality Index, scored 97.7% on Big Bench Audio and supports 97 languages with automatic language switching during conversations.
It works out of the box with LiveKit & Pipecat, no need to glue together 5 different APIs. The agent feels much more transparent since it can reason and talk simultaneously. It’s the equivalent of GPT-Live-1 but better. For now, Google is ahead of everyone else in terms of quality.
Source: deepmind.google
Free course
Stop reading about AI — start building with it
The free Claude Code course: your first site, tool or game — no coding. No upsells, no cross-sells — nothing to buy here.
Start free →▌ Related guides
Author
Evgenii Arsentev
PhD · Chief Executive Officer, digital health
Articles · Latest articles