VoxCraft — Desktop TTS Studio for Supertonic 3
Hey Supertonic team — love the v3 release. The 31-language support and expression tags are incredible.
I built a desktop studio GUI around Supertonic 3 to make it more accessible for non-developers.
Repo: https://github.com/ShyamKumar1/voxcraft
What it is:
- FastAPI backend + vanilla JS single-page frontend
- Visual voice selection (10 presets with descriptions, use cases)
- Speed slider (0.7x–2.0x), quality slider (5–12 steps), language dropdown
- WaveSurfer.js waveform preview with play/pause
- Batch mode for multiple texts, generation history with search
- Keyboard shortcuts (Cmd+Enter to generate, Space to play/pause)
- 100% local — no API calls, no data leaves the machine
Stack: Python 3.11+, FastAPI, ONNX Runtime, WaveSurfer.js. The frontend is a single HTML file — no build step, no npm.
License: MIT (code), Supertonic model uses OpenRAIL-M as required.
Would love feedback from the Supertonic community — especially on voice quality tuning, any ONNX runtime quirks to watch for, and ideas for better voice management APIs.
We have 8 "good first issue" tickets open if anyone from the Supertonic community wants to contribute.
VoxCraft — Desktop TTS Studio for Supertonic 3
Hey Supertonic team — love the v3 release. The 31-language support and expression tags are incredible.
I built a desktop studio GUI around Supertonic 3 to make it more accessible for non-developers.
Repo: https://github.com/ShyamKumar1/voxcraft
What it is:
Stack: Python 3.11+, FastAPI, ONNX Runtime, WaveSurfer.js. The frontend is a single HTML file — no build step, no npm.
License: MIT (code), Supertonic model uses OpenRAIL-M as required.
Would love feedback from the Supertonic community — especially on voice quality tuning, any ONNX runtime quirks to watch for, and ideas for better voice management APIs.
We have 8 "good first issue" tickets open if anyone from the Supertonic community wants to contribute.