Runway API

The same API generates millions of videos a month for some of the world's largest consumer technology companies. It costs $10 to start.

Start here#

Quickstart#

Install the SDK:

npm install --save @runwayml/sdk

Export your key, then generate a five-second video from a text prompt with Gen-4.5:

import RunwayML, { TaskFailedError } from '@runwayml/sdk';
 
const client = new RunwayML();
 
try {
  const task = await client.imageToVideo
    .create({
      model: 'gen4.5',
      promptText: 'A serene mountain landscape at sunrise with mist rolling through the valleys',
      ratio: '1280:720',
      duration: 5,
    })
    .waitForTaskOutput();
 
  console.log('Video URL:', task.output[0]);
} catch (error) {
  if (error instanceof TaskFailedError) {
    console.error('The video failed to generate.', error.taskDetails);
  } else {
    throw error;
  }
}

That call costs 60 credits — 12 credits per second of Gen-4.5 output, or $0.60. See pricing for every model's rate.

What you can generate#

Not sure which model to pick? A model router chooses one per request based on your cost, latency or quality preference — at no extra charge.

How the API works#

Every generation is a task. You create it, and it runs asynchronously:

1. POST /v1/image_to_video (or any generation endpoint) returns a task id immediately.
2. The task moves through PENDING → RUNNING → SUCCEEDED, FAILED or CANCELED.
3. GET /v1/tasks/{id} returns the current status, and on success an output array of URLs.

The SDKs collapse steps 2 and 3 into .waitForTaskOutput(), which polls correctly for you — with backoff, jitter and a timeout. See SDKs and task polling for when to use it and when to store the task ID instead.

Output URLs are ephemeral and expire 24–48 hours after the task completes. Download and re-host anything you show to users — see Outputs.

Scale#

Every organization starts at Tier 1: one concurrent generation, 50 per day. Tiers rise automatically as you purchase credits, with no waiting period — $100 purchased puts you on Tier 3 (5 concurrent, 1,000/day) the moment it clears.

Beyond Tier 5 there are enterprise partnerships: guaranteed concurrency, Slack support, earliest access to new models, and custom payment terms. File an exception request from the usage page of the developer portal, or contact enterprise@runwayml.com.

See usage tiers for the full table.