Press a hotkey, speak, and your words land wherever the cursor is. Cloud or fully offline with local Whisper models.
DownloadWhat's inside.
OpenAI gpt-4o-transcribe or ElevenLabs Scribe v2. Set up a priority chain — if one fails, the next kicks in automatically.
Download Whisper GGML models from HuggingFace in one click. Start a local whisper.cpp server directly from the app. Full privacy, no cloud.
System-wide shortcut that works from any app. Game-style key recorder in settings — press any combination you want.
Transcribed text is pasted directly into the focused text field via simulated Ctrl+V. No copy-paste needed.
Separate hotkey that presses Enter after pasting. Speak into Slack, Discord, or any chat — your message sends automatically.
Capture and transcribe system audio — meetings, calls, video playback. Separate hotkey for output recording.
Optional LLM pass after transcription to fix misrecognized technical terms. Uses your custom vocabulary list for corrections.
All recordings saved to disk with timestamps. Browse, replay audio, view transcriptions, re-process with different settings.
Minimal floating pill with live waveform appears on every connected display. Shows recording state without stealing focus.
Toggle between dark and light mode. System tray integration — runs in the background, launches at startup.
No installer needed. Clone the repo, run start.ps1 or npm run dev. Works on Windows, macOS, and Linux.
MIT licensed. Full source on GitHub. Inspect the code, contribute, fork it. No telemetry, no tracking.
Whisperio is free and open source. If it saves you time, consider sponsoring.