Skip to content

Repository files navigation

Foundry Vox

Foundry Vox is a macOS desktop app for local text-to-speech generation, built from PRD.md.

The current app uses a hybrid desktop architecture:

  • Tauri/Rust is the native app shell and primary app-facing API boundary
  • React/Vite renders the interface
  • Python/FastAPI remains the internal ML engine sidecar for TADA inference

The frontend no longer behaves like a generic localhost web client. Core app state, generation, clone uploads, export, setup actions, and progress updates are routed through Tauri commands.

Project layout

Verified locally

  • npm run build:web
  • uv run --project backend --python 3.12 ruff check backend/app tests/backend
  • uv run --project backend --python 3.12 python -m pytest tests/backend
  • npx tauri build --bundles app
  • npx tauri build --bundles app --config src-tauri/tauri.appstore.conf.json

Output

Runtime notes

  • Foundry Vox now prefers local model assets before attempting remote Hugging Face access.
  • On macOS, the app stores its working data under ~/Library/Application Support/Foundry Vox.
  • The model cache directory is ~/Library/Application Support/Foundry Vox/models in dev, and the packaged app exposes the same models folder through the in-app setup card.

Notes

  • The checked-in preset WAV files are development placeholders generated locally. Replace them with rights-reviewed production references before App Store submission.
  • The backend runtime is currently pinned to Python 3.11 or 3.12 because the current pydub stack is not ready for Python 3.13.
  • Actual App Store submission still requires your Apple signing identity, provisioning profile, and notarization credentials.

About

No description, website, or topics provided.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages