Getting started

The Rendley API drives the same AI video editor over plain HTTP. From your backend you can describe a video and let the agent build it, or create projects, upload media, run AI tools (transcription, image generation, video generation, translation), and render a finished MP4 yourself. No browser, no MCP client, no human in the loop.

Base URL

Every endpoint is served over HTTPS and versioned under /v1:

https://api.rendley.com/v1

Authenticate

Send your API key as a bearer token on every request. Create a key in Rendley under Settings, then API Keys. See Authentication for details.

curl https://api.rendley.com/v1/users/me \
-H "Authorization: Bearer YOUR_API_KEY"

const res = await fetch("https://api.rendley.com/v1/users/me", {
headers: { Authorization: "Bearer YOUR_API_KEY" },
});
const { data } = await res.json();

import requests

res = requests.get(
  "https://api.rendley.com/v1/users/me",
  headers={"Authorization": "Bearer YOUR_API_KEY"},
)
data = res.json()["data"]

The /templates endpoints are public; everything else requires a key.

Response envelope

All responses share one envelope. Successful calls populate data; failures populate error. meta carries pagination and other side information where relevant.

{
  "data": { },
  "error": { "code": "", "message": "" },
  "meta": { }
}

A first call to GET /users/me returns your profile, credit balance, subscription, and usage:

{
  "data": {
    "id": "usr_...",
    "name": "Ada Lovelace",
    "email": "ada@example.com",
    "credits": { "balance": 1200 },
    "subscription": { "plan_name": "Pro", "status": "active" },
    "usage": { "storage_used": 5242880000, "transcription_seconds": 480 }
  }
}

Make a video

There are two ways to produce a video over the API:

• Automated editing. Describe the video and the agent builds, edits, and exports it for you. See Automated editing.
• Build it yourself. Start from a template or assemble a timeline, then render. See Render a video.

Either way, rendering is asynchronous: you start an export, get a job_id, and poll it until the file is ready. See Jobs and polling.

The building blocks behind both:

1. Projects hold the timeline. See Projects.
2. Media comes from your uploads or from stock.
3. AI tools transcribe, generate, upscale, and translate, as async jobs. See AI tools.
4. Exports render a project to a file. See Rendering & exports.

AI tools and models

AI generation endpoints (image, video, audio, etc.) follow a two-step pattern:

1. Start the generation. Send a POST with project_id, optional model_id, and a params object. You get back a job_id.
2. Poll the job. When status is completed, result_data contains media_id and file_hash.

Every action has a default model that is used when model_id is omitted. Even though the field looks optional in the schema, a model is always selected and the default fills in automatically.

Discover available models

Each model accepts different parameters in the params object. Call GET /ai/tools to see every action, its models, and their parameter schemas:

curl https://api.rendley.com/v1/ai/tools \
  -H "Authorization: Bearer YOUR_API_KEY"

Get a download URL for generated files

AI generation jobs return media_id and file_hash, not a direct URL. To get a presigned download URL:

GET /v1/uploads?project_id=prj_123&hash=sha256_def456...

The response includes a storage_url with a time-limited presigned URL. See Jobs and polling for full examples.

Credits and errors

AI operations and exports cost credits. Every billable endpoint has a matching .../cost endpoint that returns the credit price without running the job. If your balance is too low, the API returns 402 Payment Required. See Jobs and polling for the full status-code list.

Full reference

These pages are task-focused guides. The reference pages in the sidebar cover the complete, field-by-field schema for every endpoint.