Veo 3.1
Summary
Veo 3.1 is a proprietary video generation model from Google DeepMind, released on 15 Oct 2025. It accepts text and images and generates video, with a 480-token context window and up to 8K output tokens per response.
Video model for prompt-guided generation, editing, and motion workflows
Last verified 4 Oct 2026. Every figure on this page links to its source.
Specifications
| Developer | Google DeepMind |
|---|---|
| Release date | 15 Oct 2025 [models.dev] |
| Status | Preview [models.dev] |
| Weights | Proprietary (API only) [models.dev] |
| Context window | 480 (480 tokens) [models.dev] |
| Max output | 8K tokens [models.dev] |
| Input | Text and images [models.dev] |
| Output | Video [models.dev] |
| Reasoning mode | No [models.dev] |
| Tool calling | No [models.dev] |
Veo 3.1 API pricing by provider
USD per million tokens. Sorted by input price.
| Provider | Input | Output | Cached input | Context | Source |
|---|---|---|---|---|---|
| Gemini API veo-3.1-generate-preview | – | – | – | 480 | models.dev, 4 Oct 2026 |
Frequently asked questions
What is the context window of Veo 3.1?
Veo 3.1 has a context window of 480 tokens (480) and can generate up to 8,192 tokens in a single response, according to models.dev.
Is Veo 3.1 open source?
No. Veo 3.1 is a proprietary model; Google DeepMind has not released its weights. It is available through APIs including Gemini API.
When was Veo 3.1 released?
Google DeepMind released Veo 3.1 on 15 Oct 2025, according to models.dev.
Change history
No changes detected since we started tracking this model. We re-check every source daily.