Skip to content
For developers and agents

AI video API
& MCP server.

Generate and publish faceless short-form video from your own code, workflow, or AI agent - the same pipeline the dashboard uses, with a real free test path before you spend a credit.

See Growth and Pro →Read the reference

Five pieces. All of them shipped.

Same pipeline as the dashboard
  1. 01

    REST API

    Generate a video, poll its status, and publish it to connected accounts programmatically - the same generation and publishing pipeline the dashboard uses.

  2. 02

    MCP server

    Native Model Context Protocol tools (including estimate_cost and get_channel_recommendations) so an AI agent can plan and generate video without a human driving the dashboard.

  3. 03

    Official SDKs

    @reeloop/sdk (JavaScript/TypeScript) and reeloop-sdk (Python) wrap the REST API with typed methods, kept in CI against the live API.

  4. 04

    Webhooks

    A webhook URL on generation requests gets a callback when a video finishes rendering, instead of polling.

  5. 05

    Sandbox test keys

    A dedicated sandbox key type exercises the full request path (auth, rate limiting, moderation) without spending real credits - for integration testing before going live.

FAQ

Questions,
answered.

Is there an API for generating faceless video programmatically?
Yes - a REST API generates and publishes video through the same pipeline as the dashboard, plus an MCP server for AI agents and official JS/TS and Python SDKs.
Is there an MCP server for AI video generation?
Yes. It exposes tools including cost estimation and channel recommendations, so an agent can plan a video (and its cost) before generating it.
Does the API support webhooks?
Yes - pass a webhook URL with a generation request and get a callback when the video finishes rendering, instead of polling for status.
Can I test the API without spending credits?
Yes - sandbox test keys exercise the full request path (authentication, rate limiting, moderation) without spending real credits, for integration testing before going live.
Does the API auto-publish to social platforms?
Yes - the same publishing flow the dashboard uses, to TikTok, YouTube Shorts and Instagram Reels (Facebook Reels coming soon).
Is there an AI video generator for n8n?
There is no official n8n node. n8n can call the REST API with an API key (the settings page uses "n8n workflow" as the example key name) and receive a webhook when a render finishes.
Can Zapier or Make generate a ShortLoom video?
Yes, the same way as any HTTP automation: the REST API generates and publishes, and a webhook URL on the request fires when the video is ready. There is no first-party Zapier or Make app.
Is there a Claude-compatible video generation MCP tool?
Yes. The MCP server exposes generation tools, including cost estimation, so Claude or any MCP client can plan and generate a video without driving the dashboard.
How is AI video API pricing billed per video?
The same credit rates as the dashboard. Per-video credit cost is on /video-cost-breakdown. Sandbox keys exercise the request path without spending credits.
Is there a free draft or preview step on the API?
The API and MCP default to Director Mode, a free draft. On the website, Generate spends credits immediately unless you tick Review first. The API and MCP default to a free draft, and credits move only when you approve. A Series episode is charged when it generates. The MCP tool estimate_cost exists so an agent can see the cost before generating.
Full API reference →Developer docs →Per-video cost →

Build
with it.

See Growth and Pro →
Growth and Pro · The Starter trial does not include an API key