AI Video
Generate a short video directly from an idea. Always read the catalog because capabilities differ by model.
| Model | Best for | Conversational editing |
|---|---|---|
| Gemini Omni Flash | Native audio and iterative changes | Yes |
| Veo 3.1 Fast | Faster high-quality short video | No |
| Veo 3.1 Standard | Quality-first short video | No |
Example
Use Gemini Omni Flash to create an eight-second video that moves slowly through a futuristic city at night, with natural ambient audio.
For Omni, a follow-up can say: “Keep the camera motion and make the lighting warmer.”
The Public OpenAPI currently supports text-to-video. Do not send first frames, last frames, or attachments until the public catalog exposes those capabilities through a Files API.