The Captylo app is open source.

The app’s code is on GitHub, under the GPLv3. You can check what Captylo does with your voice, build it yourself and change whatever you like.

  • Dictation, meetings, AI modes and the MCP server in one repository
  • The Pro server is a separate service that keeps no recordings and no text
  • Four commands in the Terminal and you have your own build

What you can do with the code

Captylo is free software under the GNU General Public License v3.0. You may:

  • use the app for anything you like, at work too,
  • read the code and check what happens to your voice and your text,
  • change it and build your own versions,
  • share it, with your changes too.

One condition: anything you distribute that is built on Captylo stays under the GPLv3, with its source code.

What the licence does not cover

The licence covers the code. The Captylo name, logo and app icon are not licensed for other products. If you publish a fork, please give it its own name and icon.

The animated gradient in the background (the React Bits “Grainient” shader) has its own licence, MIT with the Commons Clause, and is included under an additional permission from section 7 of the GPLv3. The details and the other components with their licences are in NOTICE.md.

Build Captylo yourself, even if you do not code

You need a Mac with Apple silicon and Xcode 26 (free in the Mac App Store, about 10 GB). An AI agent in the Terminal does the rest. Prefer to do it yourself? The commands are in the Terminal window at the top of this page.

  1. 1

    Install Xcode 26 from the Mac App Store and open it once to accept the licence.

  2. 2

    Open the Terminal and start the coding agent you use (for example Claude Code, Codex or Cursor) in your home folder.

  3. 3

    Paste the prompt next to this. The agent checks Xcode, installs the missing tools, builds Captylo and walks you through the permissions.

  4. 4

    On the first launch Captylo downloads the speech model (about 1.6 GB). After that dictation works offline.

The prompt for your AI agent
Build and install the Captylo dictation app for me from https://github.com/dawidkawalec/captylo.
1. Check that full Xcode 26 is installed (not just the Command Line Tools) and selected with xcode-select. If it is missing, stop and tell me to install it from the Mac App Store.
2. Install Homebrew if it is missing, then `brew install xcodegen`.
3. Clone the repository into ~/captylo and run `make install` there (it builds a Release version and copies Captylo.app to /Applications).
4. Open /Applications/Captylo.app and walk me through granting the Microphone and Accessibility permissions in System Settings.
If a step fails, read the error, fix it and try again. Explain what you do in simple words.

The same prompt is in the README on GitHub. A copy you build checks the same update feed as the download; to keep your own, turn off automatic checks in Settings > App.

Connect your AI assistant to your meetings

Captylo has a local, read-only MCP server: the assistant you use (for example Claude Desktop or Cursor) can search and read the meetings on your Mac. It is off by default.

  1. 1

    In Captylo open Settings > Meetings and turn on “Access for AI assistants (MCP)”.

  2. 2

    Click “Copy configuration”.

  3. 3

    Paste it into the MCP server configuration of your assistant (in Claude Desktop: Settings > Developer > Edit Config) and restart it.

  4. 4

    Ask about a meeting. The assistant starts Captylo in read-only mode by itself; the app does not need to be running.

claude_desktop_config.json
{
  "mcpServers": {
    "captylo": {
      "command": "/Applications/Captylo.app/Contents/MacOS/Captylo",
      "args": [
        "--mcp"
      ]
    }
  }
}

The path depends on where Captylo is installed. The button in the app fills in the right one.

Want to improve something? Go ahead.

Bug reports and pull requests are welcome. Start with AGENTS.md: it describes the architecture, the rules and the commands, for people and coding agents alike. A change is done when make build and make test pass.

Questions about open source

Is Captylo open source?

Yes. The app’s code is on GitHub under the GPLv3 licence: you can check what the app does with your voice, build it yourself and change whatever you like. The server behind Pro is a separate service.

See the code on GitHub

Can I build Captylo myself?

Yes. You need a Mac with Apple silicon and Xcode 26. The easiest way is to ask an AI agent in the Terminal to do it for you: the prompt is on the Open source page and in the README on GitHub. A copy you build yourself is signed only on your Mac, so macOS may ask for the permissions again after a rebuild.

Can I use the Captylo name and logo in my fork?

No. The GPLv3 covers the code, not the Captylo name, logo or app icon. If you publish your own version, give it its own name and icon.

Is the server behind Pro open too?

No. The app is open: dictation, meetings on your Mac, AI modes and the MCP server. The server behind Pro (your account, payments and Captylo AI) is a separate service. It keeps no recordings and no text, not even in its logs, as the privacy policy describes.

All questions

Or just download it.

The ready build is signed and notarized by Apple, and new versions arrive in the app. The code is the same.

Version 1.0.17 · macOS 14.4 or newer · Apple silicon · signed and checked by Apple

Source code and earlier versions