--- title: "Introducing the BlitzReels video editing API" canonical: "https://blitzreels.com/changelog/video-creation-becomes-programmable" --- # Introducing the BlitzReels video editing API Milestone: 2026-02-04 Article published: 2026-09-26 Status: Product update Create projects, upload media, edit timelines and captions, then export finished videos from your own code. Our first public API, with docs and webhooks. ![Your app, the BlitzReels API and your video linked by arrows, next to a vertical clip made through the API.](https://blitzreels.com/images/changelog/video-creation-becomes-programmable.jpg) ## When the video is one step of a bigger job Most videos start in the editor. Some start somewhere else: a course platform full of lesson recordings, a script in a spreadsheet, a tool your team built to put out a weekly recap. When the steps are the same every time, opening the dashboard for each video gets old fast. In February we introduced the BlitzReels API. It gives your code the building blocks the editor already uses, from projects and media to the timeline, captions and exports, so a repeat job can run without someone clicking through it. ## What the first version could do The API shipped with its own documentation and a full reference. With the first version, you could: - create, list and delete projects, and upload media into them - transcribe a recording, read the transcript and correct words in bulk - apply a caption style, then fix, restyle or emphasize single words - trim, split, move, transform and delete timeline items, one at a time or in batches - add text overlays, motion graphics, backgrounds and B-roll - generate a faceless video from a topic, with a voice and a visual style - start an export, follow its job and download the file A context endpoint returned the full state of a project, so an AI agent could understand it before making changes. The docs were also published as llms.txt and an OpenAPI spec, which assistants can read directly. ## How an integration fits together Every request carries an API key as a Bearer token. Whether you’re writing a one-off script or building a product, the flow looks like this: 1. Create an API key in your workspace settings (today, under API & agents). 2. Create a project in the format you need, such as 9:16 for Shorts. 3. Upload your media, or generate a faceless video into the project. 4. Apply captions and make your timeline edits. 5. Start the export, then check the job or wait for a webhook. 6. Download the file before its link expires. Long tasks like transcription, generation and exports return a job ID. Instead of polling, you can register a webhook and receive events such as `transcription.complete`, `export.ready` or `export.failed`. Each delivery is signed, so you can check it really came from BlitzReels. ## Who it’s for The API suits anyone who makes the same kind of video often enough to automate it. A few things it makes possible: - a course creator who uploads each new lesson, captions it in the channel’s style and exports a vertical teaser - a team that turns a weekly script into a narrated faceless video - a developer whose AI agent reads a project’s context, fixes caption typos and exports the result In every case you end up with a normal BlitzReels project. You can open it in the editor, check it and change it by hand before or after the export. The [timeline API](https://blitzreels.com/docs/timeline) and the [captions API](https://blitzreels.com/docs/captions) list the edits available. ## Keys, plans and limits An API key belongs to one workspace and follows that workspace’s plan. Only owners and admins can create keys, and each key is shown once, so store it somewhere safe. Rate limits and the number of jobs that can run at once depend on the plan. Storage, credits and export limits apply at the workspace level, just like in the dashboard. Today, AI generation through the API (images, videos, voiceovers and full AI videos) needs a paid plan. ## What came next Everything developers use today is built on this API. In May we added a typed TypeScript SDK and a command-line tool on top of it; here’s [how the SDK and CLI work](https://blitzreels.com/changelog/meet-the-sdk-and-cli). Today, AI assistants can also connect through our [MCP server](https://blitzreels.com/mcp), and the [AI agents page](https://blitzreels.com/agents) shows how to set up Claude Code, Codex or Cursor. To get going, the [API documentation](https://blitzreels.com/docs) has authentication, a quickstart and the full reference. ## Questions ### What can I do with the BlitzReels API? You can create projects, upload media, transcribe and caption recordings, edit the timeline, add text and graphics, and export finished videos. Long tasks return a job ID you can check, or you can receive webhook events. ### Do I need a paid plan to use the BlitzReels API? No. An API key works with any workspace and follows its plan’s rate limits, credits and export limits. Today, AI generation endpoints such as image, video and voiceover generation need a paid plan. ### Who can create a BlitzReels API key? Workspace owners and admins, from the workspace settings (under API & agents today). The key is shown once and only gives access to that workspace. ### How do I know when an export is ready? Exports run as jobs. Check the job or the export until it’s complete, or register a webhook to receive an `export.ready` or `export.failed` event. ### Can AI agents use the BlitzReels API? Yes. The docs have been published as llms.txt and an OpenAPI spec since launch, and a context endpoint gives an agent the full state of a project before it edits. Today you can also use the CLI or the MCP server. ## Build video into your own tools Read the quickstart, create a key in your workspace settings and make your first project from code. [Read the API docs](https://blitzreels.com/docs) We wrote this story later, looking back at our product history. The date at the top marks the work described here; features and access may have changed since then.