Press a global hotkey, talk, and Whisperio transcribes your speech and pastes it right where your cursor is — in any app. Cloud-fast or fully offline with local Whisper models.
Whisperio runs quietly in your tray and turns speech into text wherever you need it — with the flexibility to stay fully private.
OpenAI gpt-4o-transcribe or ElevenLabs Scribe v2. Build a priority chain — if one provider fails, the next takes over automatically.
Download GGML models in one click and run a whisper.cpp server straight from the app. Nothing leaves your machine.
A system-wide shortcut that fires from any app. A game-style recorder lets you bind any combination.
Transcribed text drops straight into the focused field via a simulated paste. No clipboard juggling.
A second hotkey presses Enter after pasting. Speak into Slack or Discord and the message just sends.
Capture and transcribe system audio — meetings, calls, playback — on its own dedicated hotkey.
An optional LLM pass fixes misheard technical terms using your custom vocabulary list.
Every recording is saved with a timestamp — browse, replay, copy, or re-transcribe with new settings.
A minimal floating pill with a live waveform shows on every display — never steals focus.
Switch themes to taste. Lives in the system tray and can launch at startup.
No installer required — clone the repo and npm run dev. Win, Mac & Linux.
Full source on GitHub under a noncommercial license. No telemetry, no tracking, no account. Inspect it, fork it, self-host it — free for any noncommercial use.
Point Whisperio at a local Whisper model and transcription happens entirely offline. Or chain providers so the cloud is just a fallback.
Run whisper.cpp, faster-whisper, LocalAI or Ollama. Works with zero internet.
Pull GGML models from HuggingFace and start a local server without leaving the app.
Prefer local, fall back to cloud — or any order you like. Whisperio retries automatically.
// First provider that succeeds wins.
// Reorder anytime in Settings → Providers.
Download, drop in an API key or a local model, set your hotkey, and talk. That's it.