https://github.com/leonburghardtdev/audio-transcriber
Offline desktop audio transcriber powered by Whisper
https://github.com/leonburghardtdev/audio-transcriber
linux macos offline pyinstaller python speech-to-text transcription tray-app whisper windows
Last synced: 27 days ago
JSON representation
Offline desktop audio transcriber powered by Whisper
- Host: GitHub
- URL: https://github.com/leonburghardtdev/audio-transcriber
- Owner: LeonBurghardtDev
- Created: 2026-06-09T14:34:06.000Z (about 1 month ago)
- Default Branch: main
- Last Pushed: 2026-06-09T16:18:46.000Z (about 1 month ago)
- Last Synced: 2026-06-09T16:25:14.293Z (about 1 month ago)
- Topics: linux, macos, offline, pyinstaller, python, speech-to-text, transcription, tray-app, whisper, windows
- Language: Python
- Size: 122 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# AudioTranscriber
> Offline speech-to-text for your desktop, powered by Whisper.
[](https://github.com/LeonBurghardtDev/Audio-Transcriber/releases/latest)
[](https://github.com/LeonBurghardtDev/Audio-Transcriber/actions/workflows/release.yml)
AudioTranscriber sits in your system tray and transcribes everything you say or hear **entirely offline** using [faster-whisper](https://github.com/SYSTRAN/faster-whisper). No cloud, no API keys, no internet required after the first model download.
---
## Features
- **Mic + Desktop audio** recorded simultaneously with labeled output
- **100% offline** — Whisper runs locally, nothing leaves your machine
- **Global shortcuts** — `Ctrl+Alt+R` toggle · `Ctrl+Alt+P` pause · `Ctrl+Alt+Q` quit
- **8 UI languages** — DE · EN · FR · ES · IT · PT · NL · PL
- **Timestamped transcripts** auto-saved to your Desktop (or custom folder)
- **Auto-start** on login, optional
- **Cross-platform** — Windows, Linux, macOS
---
## Download
Grab the latest binary from the [Releases page](https://github.com/LeonBurghardtDev/Audio-Transcriber/releases/latest):
| Platform | File |
|----------|------|
| **Windows** | `AudioTranscriber-windows.exe` |
| **Linux** | `AudioTranscriber-linux` |
| **macOS** | `AudioTranscriber-macos` |
> First run downloads the Whisper `small` model (~500 MB) once and caches it.
---
## Usage
1. Launch the app — a mic icon appears in the tray
2. **Start:** `Ctrl+Alt+R` or tray → *Start Recording*
3. **Pause/Resume:** `Ctrl+Alt+P`
4. **Stop:** `Ctrl+Alt+R` again — transcript saved, notification shown
5. **Settings:** tray → *Settings* for output folder, devices, model, language
### Tray icon states
| Color | State |
|-------|-------|
| Grey | Idle |
| Orange | Loading model |
| Red | Recording |
| Yellow | Paused |
| Blue | Transcribing |
### Output format
```
[2026-06-09 14:23:45] [Mic] Good morning, this is a test.
[2026-06-09 14:23:50] [Desktop] Welcome to the meeting.
```
---
## Build from source
```bash
git clone https://github.com/LeonBurghardtDev/Audio-Transcriber.git
cd Audio-Transcriber
pip install -r requirements.txt
python main.py
```
**Standalone binary:**
```bash
pyinstaller --onefile --noconsole \
--hidden-import soundcard --hidden-import faster_whisper \
--name AudioTranscriber main.py
```
---
## Tech stack
| Component | Library |
|-----------|---------|
| Tray icon | [pystray](https://github.com/moses-palmer/pystray) |
| Audio capture | [soundcard](https://github.com/bastibe/SoundCard) |
| Transcription | [faster-whisper](https://github.com/SYSTRAN/faster-whisper) |
| UI | tkinter (built-in) |
---
## License
MIT © LeonBurghardtDev