dictawhisper

DictaWhisper

A local voice journal. Record in the browser, drop a file, or (optionally) sync a phone folder. Transcribe on your GPU with faster-whisper. Clean the note with ollanet. The .json next to each recording is the journal.

Install from a Git checkout (below). The npm package dictawhisper 0.1.4 does not start yet: its bin runs TypeScript from node_modules, and Node refuses to strip types there. It would not install Python or faster-whisper either.

Docs: dictawhisper.com/docs · Site: dictawhisper.com

Before you clone

Exercised path: Windows, Node 22.6+ (the scripts use --experimental-strip-types), pnpm, Python with faster-whisper, and an NVIDIA GPU with CUDA. ffmpeg is required when denoise is on. macOS and Linux are unverified. CPU mode works and is slow. Cleanup via ollanet is optional.

From a checkout

git clone https://github.com/Catalyst-Forge-LLC/dictawhisper.git
cd dictawhisper
cp config.example.json config.json
pnpm install
pnpm run doctor
pnpm dev

Ready signal: pnpm run doctor exits 0, or only warnings remain. Then open http://localhost:7777. Point whisper.python at the interpreter that has CUDA Whisper. Cleanup host and model are optional; raw transcripts still work.

Quick start

pnpm dev     # API + UI  →  http://localhost:7777
pnpm start   # API only on 127.0.0.1:8008

Hit Record, or drag an audio file onto the page. Flags, HTTP, MCP, and retranscribe live in the docs.

What you get

Files are the database. Each note is the recording plus a neighboring .json. The Whisper model stays loaded. Audio and transcript files stay on this computer. Cleanup sends transcript text only to the host in ollanet.machine, and is skipped when that is empty. Search indexing embeds note text on this computer by default; it goes to another host only if you set journal.embedHost (machine, a host name, or any). With no embedding model on an allowed host, or with journal.search set to lex, nothing is embedded and search matches words only; doctor says which. The inbox is loopback by default; http.tailscale puts the same page on your tailnet. Playback follows cleaned paragraphs using Whisper word times. If cleanup is skipped or fails, the inbox shows raw only or cleanup failed, and Retry is available. Transcription itself still needs faster-whisper.

xFacts label

• AppFacts: viewer · raw
• ToolFacts: viewer · raw

Development

pnpm test
pnpm typecheck
pnpm site:dev

Site (FilePress + docs mount): pnpm --dir site ship

License

MIT · Catalyst Forge LLC

See the rest of the Catalyst Forge shelf.