AI video generation API
Send a prompt, get back a finished video: Yoh writes the script, casts consistent characters, adds voices, music and captions and renders the MP4. Or drive each step yourself: stories, series, clips, characters, timeline and renders.
Your first video in three calls
Create a key
Sign in, open Settings → API keys and create a key with stories:generate and stories:read. It is shown once; keep it in a server-side secret. Check it:
curl https://www.yoh.app/api/v1/credits \ -H "Authorization: Bearer $YOH_API_KEY"Generate
X-Max-Generation-Credits is your consent: the most credits you accept for this request. A higher price returns HTTP 409 with the quote and charges nothing. Reuse the same Idempotency-Key to retry safely.
curl -X POST https://www.yoh.app/api/v1/stories/generate \ -H "Authorization: Bearer $YOH_API_KEY" \ -H "X-Max-Generation-Credits: 500" \ -H "Content-Type: application/json" \ -H "Idempotency-Key: my-first-video-1" \ -d '{ "prompt": "A dragon teaches a young knight how to fly", "contentType": "cinematic_story", "aspectRatio": "16:9", "scope": "full", "renderVideo": true }'Poll the job
Poll every 20 seconds until the status is completed; the answer carries the storyId and the videoUrl of the MP4.
curl https://www.yoh.app/api/v1/stories/jobs/JOB_ID \ -H "Authorization: Bearer $YOH_API_KEY"
Why it works well
Full control when you want it
Create a series and its cast once, then make episodes in it; edit clips, regenerate one clip's picture or video, start from your own uploaded video, and render the timeline you choose.
Spending you can predict
Every paid request needs a credit cap or an accepted quote. Free previews (estimate, quoteOnly, dryRun) tell you the price before anything starts.
Built for retries
Idempotency keys on generation, durable jobs you can poll and cancel, and clear terminal states.
Prefer an agent?
The same account and scopes work over MCP, so Claude or Codex can call Yoh for you.
Questions
Is there an API to generate AI videos from text?
Yes. POST /api/v1/stories/generate takes a prompt and returns a job; when it completes you get a rendered MP4 with script, characters, voices, music and captions.
How is the API priced?
It uses the same Yoh credits as the app. Each paid request states the most credits you accept, and you can get the exact price first from the free estimate endpoints.
Can I keep the same characters across videos?
Yes. Create a series with its characters once, then generate episodes inside it; every episode reuses the same cast.
Where is the full reference?
The docs list every endpoint, scope and error with a live playground.
Full reference: Yoh docs · Pricing · OpenAPI 3.1
Other ways to connect
- Make AI videos with ClaudeConnect Claude to your Yoh account and make video stories by asking. Claude writes the script, keeps characters consistent, generates the clips, edits the timeline and renders the finished video, and checks with you before anything costs credits.
- Make AI videos with CodexConnect Codex to Yoh and let it produce video stories from your terminal or editor: script, cast, clips, voices, captions and the final render, with a price shown and approved before every paid step.
- The Yoh MCP server for AI videoOne remote MCP endpoint gives any agent the whole Yoh studio: write a story, cast and draw characters, generate clip videos with voices and music, edit the timeline and captions, render and publish. Every tool follows the same rules as the yoh.app editor.
- The Yoh skill for AI agentsThe MCP server gives your agent the tools; the skill teaches it the craft. A SKILL.md in the open Agent Skills format tells the agent how to make a Yoh story well and the rules that are easy to get wrong.