Overview
The real-time audio-to-audio member of Gemini 3.8 for low-latency voice agents and natural live dialogue.
Gemini 3.8 Live
The real-time audio-to-audio member of Gemini 3.8 for low-latency voice agents and natural live dialogue.
Google announced Gemini 3.8 Live GA in the Live API on 2026-09-15; it is the default option for low-latency voice agents and real-time dialogue, with no shutdown date announced.
MODEL NUMBER080
- First public date
- 2026-09-15
- Organization
- Google DeepMind
- Family
- Gemini
01
Specifications
Only published specifications are shown.
- Input modalities
- Text · Audio
- Output modalities
- Audio
- Open weights
- No
- Status
- Current
02
Overview
Why it mattered
It brings Gemini 3.8 reasoning, asynchronous tool calling, and full session content updates into the production Live API as Google’s stable real-time voice line.
03
Architecture & capabilities
Model capabilities
Native audioReal-time audioThinking modeTool use
04
Release history
Gemini 3.8 Live reached GA in the Live API as the default audio-to-audio model for low-latency voice agents and real-time dialogue.
View source05
Related models
Select a card to view the related model.
06
Sources
Suggest a correction
Copy this template, add the field, proposed value, and primary source, and share it with the archive maintainer.