Gemini 3.1 Flash Live Preview
Google's low-latency, audio-to-audio model optimized for real-time dialogue and voice-first AI applications, with acoustic nuance detection, numeric precision, and multimodal awareness.
text + image + audio + video → text + audio · made by Google
Free
Offered at no charge, which is not the same as cheap: a free lane carries a rate limit and can be withdrawn. The prices above are what you pay when it is not available to you.
Sold by 2 ways
| Seller | Lane | Rate |
|---|---|---|
| Google api | free | $0per Mtok in$0per Mtok outai.google.dev · read 2026-09-16 |
| Google 0 ways | $0.75per Mtok in$4.5per Mtok out | |
| Google apinot on the seller's list today | standard | $0.75per Mtok in$4.5per Mtok outai.google.dev · read 2026-08-24 |
Measured 2 standings
| Place | Board | Metric | Score |
|---|---|---|---|
| 8thof 16 | τ³-Voice Leaderboard | Pass^1 | 43.8 |
| 8thof 16 | τ³-Voice Leaderboard | Pass^1 (high) | 43.8 |
About
Gemini 3.1 Flash Live Preview — a text and audio model from Google, placed 8th of 16 on τ³-Voice Leaderboard.
It takes text, images, audio and video and returns text and audio, with a context window of 131,072 tokens. It was published in March 2026, trained on material up to January 2025. Its sellers say it can reason step by step and call a tool. The catalogue files it under speak. It stands 8th of 16 on τ³-Voice Leaderboard.
Every current figure
- Maker
- Register
- model
- Takes
- text + image + audio + video
- Returns
- text + audio
- Context
- 131,072 tokens
- Longest answer
- 65,536 tokens
- Published
- March 2026
- Knowledge to
- January 2025
- Licence
- not read
- Boards
- 1
- Best place
- 8th of 16 — τ³-Voice Leaderboard
Known as 3 names
gemini-3.1-flash-live-preview-thinking-highgemini-3.1-flash-live-preview-thinking-minimalgemini-3.1-flash-live-preview