Metadata-Version: 2.4
Name: blurt
Version: 0.3.9
Summary: Talk, don't type. Runs on-device on macOS Apple Silicon. No cloud, no API keys and forever free.
Author: Satya Borgohain
License-Expression: MIT
Project-URL: Homepage, https://github.com/satyaborg/blurt
Keywords: speech-to-text,whisper,mlx,apple-silicon,macos,voice,transcription
Classifier: Development Status :: 4 - Beta
Classifier: Environment :: MacOS X
Classifier: Intended Audience :: End Users/Desktop
Classifier: Operating System :: MacOS
Classifier: Programming Language :: Python :: 3
Classifier: Topic :: Multimedia :: Sound/Audio :: Speech
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: mlx-whisper
Requires-Dist: sounddevice
Requires-Dist: pynput
Requires-Dist: numpy
Requires-Dist: rich
Provides-Extra: dev
Requires-Dist: pytest; extra == "dev"
Requires-Dist: ruff; extra == "dev"
Dynamic: license-file

# Blurt

Talk, don't type.

![demo](https://raw.githubusercontent.com/satyaborg/blurt/main/demo.gif)

Hold the right cmd ⌘, speak and release - text appears wherever your cursor is. Runs on-device on macOS Apple Silicon.

## Requirements

- macOS with Apple Silicon (M1/M2/M3/M4)
- Python 3.10+

## Install

```bash
pipx install blurt
```

Requires [pipx](https://pipx.pypa.io/) (`brew install pipx`).

Update to latest: `pipx upgrade blurt`

First run downloads the Whisper model (~1.6 GB).

macOS will prompt you to grant your terminal:
- **Microphone** access
- **Accessibility** access (System Settings → Privacy & Security)

## Vocab

Add words STT models often get wrong (names, jargon, acronyms):

```bash
blurt vocab add Claude Code   # add
blurt vocab                   # list
blurt vocab rm Claude Code    # remove
```

Or edit manually in `~/.blurt/vocab.txt`.

## Contributing

```bash
git clone https://github.com/satyaborg/blurt.git
cd blurt
pipx install . --force
```

This installs the local version of blurt. After making changes, re-run `pipx install . --force` to test them.

## Privacy

Your audio never leaves your Mac. Everything including recording, transcription and inference runs locally. No network calls, no telemetry, no accounts, and no API keys needed.

## License

MIT
