Gemini 3.8 Live Extended Thinking
by Vercel AI Gateway · family gemini-flash · listed Sep 2026T
Live API voice model with configurable background reasoning for multi-step real-time agents
Current rates USD per 1M tokens
Input$0.75
Output$4.50
Cache read—
Cache write—
Against the market
Input price vs. maker flagships ($/1M tokens)Output price vs. maker flagships ($/1M tokens)On the record
- Model id
google/gemini-3.8-live-extended-thinking- Context window
- 131K tokens
- Max output
- 66K tokens
- Input modalities
- text, audio
- Output modalities
- text, audio
- Reasoning
- no
- Tool calls
- yes
- Open weights
- no
- Knowledge cutoff
- —
- Released
- 2026-09-15
- Last updated
- 2026-09-15
← All Vercel AI Gateway listings