Which Google Gemini models create your videos and check your conditions.
VEOit uses Google's Gemini API for all AI work. You do not need your own Google account or API key.
Video models
- Veo 3.1 Lite: the main model. It turns your first product image into an 8-second, 720p, 16:9 video
- Veo 3.1 Fast: the fallback model, used if the main generation fails. It can also work from up to 3 reference images at once
Condition model
A fast Gemini text model reads the product title, description and image count to answer your Video Generation Condition with yes or no.
Image-to-video
Because the video starts from your real product photo, the product keeps its shape, colors and details. The model adds motion, lighting and a setting around it based on your instructions.