[Shotstack](https://shotstack.io/) is a video editor that runs in the cloud and is driven by software. AI agents,
automations and apps send an edit, and Shotstack returns the finished video.

An edit is a JSON description of a video: the clips, text, audio and effects on a timeline, and the output you want. You
can write it by hand, build it in your code, or ask an AI agent to write it for you.

## One edit, everywhere

The same edit moves between every part of Shotstack without conversion:

- The **Edit API** renders it in the cloud and returns a video, image or audio file.
- **Studio** opens it on a visual timeline, so a person can review and change it.
- A **template** saves it with merge fields such as `{{NAME}}`, so you can render it again with new data.

An agent can draft a video, a person can adjust it in Studio, and your app can render a version of it for every record in
your data.

## Three ways in

| Way in                | For                                         | How                                                                                                                                           |
| :-------------------- | :------------------------------------------ | :-------------------------------------------------------------------------------------------------------------------------------------------- |
| With an AI agent      | Anyone using Claude, ChatGPT, Cursor or similar | The [MCP server](/agents/mcp-server) for chat clients and IDEs, the [CLI and skill](/agents/cli) for coding agents, or [Director](/agents/director) in Studio |
| In code and workflows | Developers and automation builders          | The [Edit API](/getting-started/hello-world-using-curl) and SDKs from your backend, or the [n8n, Make and Zapier apps](/architecting-an-application/automation-tools)                                   |
| Inside your product   | Product teams                               | The [Studio SDK](/studio-sdk), a video editor you embed in your app                                                                |

## What the engine does

- **Renders** edits in the cloud, in parallel, and tells you by webhook when each one is ready.
- **Generates** video, images, voice-overs and music from prompts while the video renders.
- **Ingests** your source files, transcodes them and transcribes speech with the Ingest API.
- **Serves** finished files from a CDN, or delivers them to your own storage.

## Start here

- [Make your first video with an AI agent](/getting-started/quickstart-agent), with no API key needed to preview.
- [Make your first video with the API](/getting-started/hello-world-using-curl), using curl.
- [Learn how an edit is structured](/getting-started/core-concepts).