OS Voice v0.1.0

OS Voice v0.1.0

Local-first, AI-powered voice-to-text that types into any app — private by design.
A product of OSA · by Miosa.ai.

Press a hotkey, speak, and your words are transcribed locally (GPU-accelerated) and typed into any app — optionally cleaned up into structured text by a local AI. Runs 100% offline: no account, no cloud, no telemetry.

Downloads

Platform	File
🍎 macOS (Apple Silicon + Intel)	OS Voice_0.1.0_universal.dmg
🪟 Windows	OS Voice_0.1.0_x64-setup.exe (or the .msi)
🐧 Linux (Debian / Ubuntu)	OS Voice_0.1.0_amd64.deb

Install notes

These builds are unsigned (no paid Apple/Microsoft signing certificate), so:

• macOS: open the .dmg, drag to Applications, then right-click the app → Open → Open anyway (or run xattr -cr "/Applications/OS Voice.app"). Grant Microphone + Accessibility permissions on first launch.
• Windows: if SmartScreen warns, click More info → Run anyway.
• Linux: sudo apt install ./OS\ Voice_0.1.0_amd64.deb

Optional: AI cleanup

Transcription works fully offline on its own. For smart formatting (filler removal, punctuation, topic paragraphs), install Ollama and pull a model — e.g. ollama pull qwen2.5:14b — then enable it in Settings → AI Post-Processing. On Apple Silicon and modern GPUs it runs locally in ~1s.

Credits

OS Voice is a rebrand and continuation of the excellent open-source Voquill (AGPLv3) by Voquill, Inc. Huge thanks to the Voquill team — this wouldn't exist without their work. 🙏

Licensed under AGPLv3.