Offline medical dictation app. Record, transcribe, and generate clinical notes -- all locally. Patient data never leaves your device.

• Speech-to-text -- transcribe medical dictation using a locally-running speech recognition model
• AI-powered notes -- optionally generate corrected transcripts and SOAP notes using a local language model
• Fully offline -- everything runs on your device, so patient data never leaves the machine
• Pluggable models -- swap in different speech recognition or language models as they become available

Download the latest release for your platform from the Releases page:

• macOS -- .dmg
• Windows -- .msi or .exe

After installing, you'll need to download a speech recognition model:

1. Go to the plugins page and download a Dictation plugin.
2. Open Dr. Toru, go to Settings, and import the downloaded file.

To generate corrected transcripts and SOAP notes, import a .llamafile language model from the same Settings screen.

• pnpm 10+
• Tauri 2 system dependencies (macOS / Windows)

pnpm install
pnpm tauri dev

pnpm tauri build

pnpm test            # unit tests (vitest)
pnpm check           # typecheck + lint + format check

src/                  # TypeScript frontend
  app/                #   dictation controller, recording service, UI
  plugins/            #   plugin contracts, registry, runtime adapters
  asr.worker.ts       #   ASR inference (Web Worker)
src-tauri/            # Rust backend
  src/plugins/        #   plugin storage, import, llamafile runtime
  src/storage.rs      #   recording persistence