Gemini 3.5 Transcribe
Summary
Gemini 3.5 Transcribe is a proprietary language model from Google DeepMind, released on 26 Aug 2026. It accepts audio and generates text, with a 98K-token context window and up to 33K output tokens per response. As of 21 Oct 2026, the lowest listed API price is $2 per million input tokens and $12 per million output tokens (Vercel AI Gateway), across 1 provider we track.
Speech-to-text model with language detection, speaker diarization, word timestamps, and custom vocabulary
Last verified 21 Oct 2026. Every figure on this page links to its source.
Specifications
| Developer | Google DeepMind |
|---|---|
| Release date | 26 Aug 2026 [models.dev] |
| Status | Generally available [models.dev] |
| Weights | Proprietary (API only) [models.dev] |
| Context window | 98K (98,304 tokens) [models.dev] |
| Max output | 33K tokens [models.dev] |
| Input | Audio [models.dev] |
| Output | Text [models.dev] |
| Reasoning mode | No [models.dev] |
| Tool calling | No [models.dev] |
Gemini 3.5 Transcribe API pricing by provider
USD per million tokens. Sorted by input price.
| Provider | Input | Output | Cached input | Context | Source |
|---|---|---|---|---|---|
| Vercel AI Gateway google/gemini-3.5-transcribe | $2 | $12 | – | 98K | models.dev, 21 Oct 2026 |
Frequently asked questions
What is the context window of Gemini 3.5 Transcribe?
Gemini 3.5 Transcribe has a context window of 98,304 tokens (98K) and can generate up to 32,768 tokens in a single response, according to models.dev.
How much does the Gemini 3.5 Transcribe API cost?
Gemini 3.5 Transcribe is listed from $2 per million input tokens and $12 per million output tokens (Vercel AI Gateway). Prices last verified 21 Oct 2026.
Is Gemini 3.5 Transcribe open source?
No. Gemini 3.5 Transcribe is a proprietary model; Google DeepMind has not released its weights. It is available through APIs including Vercel AI Gateway.
When was Gemini 3.5 Transcribe released?
Google DeepMind released Gemini 3.5 Transcribe on 26 Aug 2026, according to models.dev.
Change history
- Context window changed from none to 98K
- Max output changed from none to 33K
- Added to the directory
- Now listed on Vercel AI Gateway