Overview

Video generation in AI Studio turns a start image (and optional end image or prompt) into short AI videos. Generation is queued, uses your credit balance, and results stay cached per workspace for a day.

Opening video generation

1. Go to AI Studio and switch Mode to Video
2. Or open Create > Video to use the dedicated video creation page
3. Your recent results appear in a grid and are cached for 24 hours per workspace

Generating a video

1. Add a start image (required) and optionally an end image and a text prompt to guide the motion
2. Choose a model: Minimax, Runway, Pika, or Google Veo
3. Set the duration and any additional model settings
4. Click Generate. Your request is added to the queue
5. The status updates automatically from queued to processing to completed. You can keep browsing while it runs
6. When completed, the video appears in the results grid and is ready to preview

Managing results

1. Click a video to preview it
2. Select one or more videos and use the bulk actions bar to save them to your media library
3. Clear results with the Clear button in the header
4. Cached videos for the current workspace restore automatically for 24 hours when you return

Credits and queue processing

• Each generation reserves credits before the provider is called. If credits are insufficient, the request will not start.
• Jobs move through queued, processing, and then completed or failed. If a job fails, an error message is shown and you can try again.

Using a generated video in a post

1. Select a video and click Add to post or save it to your library first and attach it from there
2. Insert it into the composer tab for the target platform
3. Schedule or publish as usual

Related articles

• Draft the surrounding caption in Composer
• See how video posts appear on the schedule in Calendar