Welcome

Video generation is still in an experimental phase. Thank you for being among the first to try it on OpenRouter.

What we need from you:

• Try the API with realistic workloads. We support text-to-video and image-to-video.
• Tell us what breaks, confuses, or feels wrong about the API. We'll lock it down at the v1 launch, so now is the time to speak up.

We expect this experimental period to run 1-2 weeks. Everything here, including the API spec, is subject to change.

Questions & Feedback

We’re eager to hear what you think! Drop any feedback, positive or negative, in #video-feedback on Discord. If you haven’t signed up for our Discord yet, go here.

We’d prefer to use Discord to share all feedback, but if that’s not an option, you can email feedback to [email protected]

Known Limitations

• No SDK support (coming at v1 launch)
• Model pages and pricing displays on the website are not fully available
• No cancel or delete endpoints

Billing & Credits

Billing is live during the alpha so we can test the full payment flow. If a generation fails or behaves unexpectedly, drop a note in #video-feedback on Discord and we'll credit you back. We also have promo credits set aside for testers who help us find and fix issues.

When you submit a generation, credits are placed on hold based on an estimated cost. When the generation is complete, the hold is removed, and the actual charge is applied. If the generation fails, the hold is released.

You may see held credits affect your available balance for other API calls. This is expected and you can see all the generations on the Jobs tabs on the Logs page.

Available Models

Our docs are still in progress, so you’ll need to reference the models providers for details. You can also use the /api/alpha/videos/models endpoint to programmatically retrieve the accepted parameters for each model.

Google Veo 3.1 (google/veo-3.1)