AI video generation API

An AI video generation API that returns a finished video, not a clip

Most video APIs give you a few seconds of footage and leave the rest to you: the script, the voice, the edit, the captions, the vertical cut. The Chitrio API does the whole job. Send one POST with a topic or a brief, and Chitrio writes the script, voices it, builds every scene with AI motion, burns in captions, adds music and returns a web-ready 16:9 MP4, plus a vertical Short if you ask for one. It is asynchronous: you get a jobId at once, then a signed webhook when the video is ready, or you poll. Authentication is one workspace API key. The same engine powers the Chitrio app, so a video made through the API looks exactly like one made by hand, with the workspace voice, caption style and brand colour applied.

Free trial to start · the cost is shown before every render

Straight answer: This is a production API for finished videos, not a raw model endpoint. You do not pick a model or tune frames; you describe the video and get the file. Renders take a few minutes, so design for a webhook or a poll, not a blocking request.

How it works

  1. 1

    Create a key

    Sign in, open Settings, then API keys, and create a key. It is shown once and acts on one workspace, its balance and its settings.

  2. 2

    Start a video

    POST /api/v1/videos with a topic, a length of 1 to 20 minutes, a language, a tone and a visual style. You get 202 and a jobId.

  3. 3

    Get the file

    Receive a signed video.completed webhook or poll GET /api/v1/videos/{jobId}, then download the MP4 and the optional Short.

The API call

Real endpoints and fields. The full reference, with Node.js and Python examples, is on the developers page.

Request: start a video

curl -X POST https://chitrio.com/api/v1/videos \
  -H "X-Api-Key: YOUR_API_KEY" -H "Content-Type: application/json" \
  -d '{"topic":"How the world's fastest trains work","durationInMinutes":3,"language":"en","generateShort":true,"webhookUrl":"https://your-server.com/chitrio/webhook"}'

Response

HTTP/1.1 202 Accepted
{ "jobId": "3f1c9a0e-...", "status": "Processing", "title": "How the World's Fastest Trains Work" }

Status when it is done

curl https://chitrio.com/api/v1/videos/3f1c9a0e-... -H "X-Api-Key: YOUR_API_KEY"

{
  "jobId": "3f1c9a0e-...",
  "status": "Completed",
  "progress": 100,
  "videoUrl": "https://chitrio.com/api/v1/videos/3f1c9a0e-.../download",
  "shortUrl": "https://chitrio.com/api/v1/videos/3f1c9a0e-.../download?variant=short",
  "error": null
}

Chitrio features used

Named as they appear in the product, so you can find each one after you sign in.

  • The API: video inside any product

    An LMS, a CRM or an HR platform sends a topic and gets a finished video back.

  • Shorts from every video

    Make a Short from one idea, or get a vertical Short and free Status cuts from every finished video.

What people make with it

Video inside your own product

Let your users type a subject in your app and get a narrated video back, without building a video pipeline.

Content pipelines

Turn a queue of topics, articles or product names into videos overnight, three renders at a time per workspace.

Learning platforms

Generate explainer videos for lessons in English, Hindi and other render languages, with captions in the right script.

Social schedulers

Ask for generateShort and receive a 9:16 cut with captions for Shorts, Reels and Status alongside the main video.

Internal tools

Wire the API into a no-code tool or a cron job: it is plain JSON over HTTPS with one header for auth.

Questions, answered

What do I get back?

A finished 16:9 MP4 with narration, moving scenes, captions and music, and with generateShort a vertical Short of up to 60 seconds. Both are downloaded with your API key.

How long does a render take?

A short video usually takes a few minutes; long videos and busy periods take longer. Use the signed webhook in production, or poll every 15 to 30 seconds.

How is it billed?

Every render uses the workspace credits, priced the same way as in the app. GET /api/v1/estimate quotes a video before you start, and credits are returned automatically if a render fails or is cancelled.

Which plans include the API?

Any workspace can create API keys. What you can render depends on the credit balance and plan of that workspace, exactly as in the app.

Are there rate limits?

Starting work is limited to 20 requests per 10 minutes per key, and up to 3 videos of a workspace render at once. A 429 tells you when to retry.

Related pages

Send your first topic today

Create a workspace, add a key under Settings and make one call. The full reference is on the developers page.

Create a workspace and a key