Skip to main content

How to use the Vimeo AI API

—transcribe, translate, dub, and ask questions via REST API endpoints, without using the Vimeo web interface.

In this article:

Who can use the Vimeo AI API?

The Vimeo AI API is available to Enterprise plan users only. Contact your account manager if you'd like to upgrade.

Available actions

The API supports four actions under /videos/{id}/ai/*:

  • Transcription — Generate a transcript from your video's audio
  • Subtitle Translation — Translate subtitles into supported languages
  • Audio Dubbing — Dub your video's audio into another language
  • AskAI — Ask questions about your video's content and get AI-generated answers

There are nine endpoints in total: POST and GET for each action, plus a GET endpoint for supported languages.

How it works

All AI API calls are asynchronous:

  1. Send a POST request to start a job (e.g., /videos/{id}/ai/transcription)
  2. Poll the status with a GET request to the same endpoint
  3. When the job completes, your result is ready

Each call targets a single video. There's no bulk endpoint—to process multiple videos, loop through them individually.

Authentication: the ai OAuth scope

To call AI API endpoints, your token needs the new ai scope. You can add it when generating a personal access token—no provisioning or manual setup required.

AI credits

API requests draw from the same AI credit pool as the Vimeo web interface, at the same rates. For example, subtitle translation still costs 10 credits per minute of video per language. See AI credits rates and policies for full details.

Rate limits

Each AI API endpoint is limited to 10 requests per minute (~600 videos per hour). If you need a higher limit, your account manager can adjust it for your account.

API documentation

Full endpoint specs, request/response formats, and examples are at developer.vimeo.com/api/reference/videos#ai.

Suggested next steps

Related articles