Overview
Veo 3.1 Preview is Google’s preview video-generation model for creating short clips from text and images, with synchronized audio. It was announced for Vertex AI on October 15, 2025. The standard preview is distinct from Veo 3.1 Fast Preview and the later generally available model. (docs.cloud.google.com)
Capabilities and uses
- Generate video from a text prompt or an image; guide generation using reference images or first and last frames.
- Generate synchronized dialogue and sound effects. Google’s guide describes 4-, 6-, and 8-second clips, 16:9 and 9:16 aspect ratios, and 720p and 1080p output; feature availability and parameter combinations depend on the service and configuration.
- Suitable for creative prototyping, storytelling, advertising concepts, and integrating generated video into applications. (cloud.google.com)
Limitations and availability
This was a preview endpoint, not the generally available endpoint. On Vertex AI, Google listed it as discontinued and recommended migrating to veo-3.1-generate-001 before April 2, 2026. In the Gemini API, Google states that it was shut down on October 22, 2026, and recommends gemini-omni-1.2-flash or a generally available Veo model on Google Cloud. These are platform-specific status notices; the replacement IDs are different models or endpoints. (docs.cloud.google.com)
Pricing
Checked November 17, 2026. Current pricing for this discontinued preview endpoint is not verified. A previously indexed Google Gemini Developer API pricing page listed the standard model’s video-with-audio output at USD $0.40 per generated second for 720p or 1080p and USD $0.60 per generated second for 4K; it did not state a separate input-token charge in the cited table. That indexed listing is historical and should not be treated as a current offer. The current pricing page no longer provides a verifiable price for this endpoint. (ai.google.dev)