Docs

This is the product docs hub. Full technical detail lives in the main repository README and source. This page collects the paths people use most, with links to guides and posts.

Quick start

1. Install the desktop app — Use the site download page, or grab macOS (Apple Silicon / Intel) and Windows builds from GitHub Releases. Early builds may be unsigned; approve them in system settings on first launch.
2. Or run from source — Requires Node.js 24.x and npm. See the repo README and Run OpenChatCut from source.
3. First conversational edit — Import media, describe the goal in chat (pacing, captions, music, aspect ratio), let the agent change the timeline, then review and refine in the preview.

Core topics

Install and desktop builds

• Desktop install and common issues
• GitHub Releases
• Main repo README: Quick Start

First conversational edit

• What is OpenChatCut
• AI agent video editing: from prompt to editable timeline
• Homepage workflow: /#workflow

MCP and external agents

OpenChatCut serves a local Streamable HTTP MCP endpoint (default):

http://localhost:5199/api/external-mcp/mcp

• Connect Codex / Claude Code with MCP
• Main repo notes: README

Captions and export

• Caption workflow: transcription, style, and export
• Export video, audio, captions, and project formats such as FCPXML (capabilities depend on the current release)
• Watch frame range, resolution, and progress feedback during export

Models and keys

• Provider keys stay on the local server side; the browser does not hold secrets
• The built-in agent supports multiple providers and OpenAI-compatible APIs
• See in-app settings and the repository docs for details

Releases and philosophy

• Changelog
• About
• Why not pure AI video generation

Community

• Discord
• GitHub Issues
• GitHub Discussions (when enabled)

Legal

• Privacy Policy
• Terms of Service

Docs evolve with releases. If a link here diverges from the repository, trust the OpenChatCut main repo.