Google Releases Gemini 3.8 Live and 3.8 Live Extended Thinking for Production Grade Voice Agents
Google's Gemini 3.8 Live Extended Thinking hits 82.6 on speech-quality benchmarks — and it's in production today at $0.005/min.
Why it matters
Google ships a production-grade voice model with extended reasoning and tool calling, positioning itself for the agent-native era. Practitioners can deploy multimodal, multilingual voice agents immediately; enthusiasts see another major lab capability release in what's becoming a weekly cadence of frontier model drops.
The key facts
9 to knowGemini 3.8 Live and 3.8 Live Extended Thinking released to production
Extended Thinking ranks #1 on Artificial Analysis' Speech to Speech Quality Index with 82.6
Scores 97.7% on Big Bench Audio benchmark
Supports 97 languages with mid-conversation switching
Tool and API execution in background during conversation
Live visual input processing
Pricing: $0.005/min for audio input
Available in Gemini API and Google AI Studio
Audio output watermarked with Google DeepMind's SynthID
Go to the source
MarkTechPostmarktechpost.com
Publisher excerpt: Google has released Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, its most advanced live dialogue models to date. The models execute tools and API calls in the background while the conversation keeps flowing, process live visual inputs, and switch between 97 languages mid conversation.…