Privacy-first, offline desktop translation app for macOS. All translation runs locally on your machine using TranslateGemma models accelerated by Apple Silicon — no data ever leaves your device.
- Fully offline — no internet required after initial model download
- 79 languages — powered by Google's TranslateGemma models
- Multiple models — choose between speed (4B) and accuracy (27B)
- Text-to-speech — optional Qwen3 TTS for 10 languages
- GPU accelerated — runs on Apple Silicon via MLX
brew install fcjr/fcjr/local-translateDownload the latest DMG from Releases.
| Model | RAM | Download Size |
|---|---|---|
| TranslateGemma 4B (4-bit) | 4 GB | ~2.2 GB |
| TranslateGemma 4B (8-bit) | 6 GB | ~4.1 GB |
| TranslateGemma 27B (4-bit) | 18 GB | ~15.2 GB |
Models are downloaded on first use and cached in the Hugging Face Hub cache (~/.cache/huggingface/hub/).
- macOS Ventura (13.0) or later
- Apple Silicon (M1 or later)
pnpm install
uv sync
pnpm tauri devRequires Node.js, Rust, and Python 3.10-3.12.