← Back to Voice models
Google DeepMindNotable
Gemini 3.1 Flash Live
Released Mar 26, 2026
Google's earlier audio-to-audio Live API model, still in preview. Google now labels it legacy and recommends Gemini 3.8 Live.
- Price per minute
- —
- Pricing
- Shares the Gemini 3.8 Live table: input text $0.75 / audio $3.00 ($0.005/min) / image-video $1.00 ($0.002/min); output text $4.50 / audio $12.00 ($0.018/min) per 1M tokens.
- Inputs
- audio, text, image
- Tool calling
- Yes
- Languages
- —
- Voices
- 30
- Big Bench Audio
- 96.6%
- Access
- Beta
- Licence
- Proprietary
Strengths & weaknesses
Strengths
- Big Bench Audio 96.6% at high thinking (AA)
- Thinking levels from minimal to high
- Same per-token price as Gemini 3.8 Live
Weaknesses
- Legacy preview: Google recommends migrating to Gemini 3.8 Live
- Synchronous function calling only
- 2.99s time to first audio at high thinking (AA)
Notable: frontier-lab clause. Released 2026-03-26 per the Gemini API changelog. The model page calls it a 'legacy preview audio-to-audio model', with no deprecation date; drop it if legacy models are out of scope. AA: High 96.6% (S2S 72.0), Minimal 71.3% (S2S 64.2).
Links
API model strings
- google
gemini-3.1-flash-live-preview