Press a hotkey, speak, and your words appear at your cursor, in any app. Transcription runs on your own computer. No cloud, no account, no tracking.
Runs on macOS, Windows and Linux. .deb and all builds on GitHub
An email, a chat, a whole chapter, a messy thought and a two hour lecture. Watch Wspr type all of it.
From any app. Tap to start and stop, or hold it down and let go when you're done.
A small pill shows you're recording. Talk the way you would to a colleague.
Your text lands at the cursor, or on the clipboard if there's nowhere to type.
Speech becomes text on your own machine, with the GPU you already own. Nothing is uploaded, so there's nothing to review, leak or sell.
Turn on Rewrite and Wspr tidies the ums, the repeats and the half sentences before the text lands. Pick a style, or write your own.
um, so like i think the meeting uh is at three 3 PM tomorrow tomorrow tomorrow you know
The meeting is at 3 PM tomorrow.
Drop in a lecture, an interview or a voice memo. MP3, WAV, M4A, MP4, MOV, MKV and more. A two hour recording never blocks your hotkey.
Detected automatically. Wspr writes in the language you speak.
Hold the hotkey, talk, let go. Ctrl+Shift+Space by default, yours to change.
Every transcript kept, and stored only on your machine.
Hint the names and jargon you use, so they come out spelled right.
The same app, the same features, on the computer you already have.
No window to manage. Wspr waits in the menu bar or system tray until you press the key, and plays a soft tone when it starts listening.
One idea per short. Made to be shared.
No subscription, no renewals. Start free, and pay once when it earns it.
Everything you need to try it properly.
About what one month of a dictation subscription costs.
| Feature | Wspr | Wispr Flow | Superwhisper | MacWhisper |
|---|---|---|---|---|
| Price | $14.99 lifetime | $12 to $15 / month | Subscription or lifetime | $29 to $80 |
| Platforms | macOS, Windows, Linux | macOS, Windows, iOS, Android | macOS, Windows, iOS | macOS |
| On-device transcription | ✓ | Cloud | ✓ | ✓ |
| AI rewrite with your own key | OpenAI, Anthropic, Gemini | Proprietary | Multiple (your key) | OpenAI, Claude |
| Live, system-wide dictation | ✓ | ✓ | ✓ | ✓ |
| File transcription | ✓ | Not advertised | ✓ | ✓ |
| One-time purchase available | ✓ | ✗ | ✓ (lifetime tier) | ✓ |
Yes. 50 live transcriptions, 5 file transcriptions and the built-in rewrite styles, with no card and no account. Pro unlocks everything for a single $14.99, paid once.
Always. Transcription runs locally with whisper.cpp on your GPU, and your audio never leaves your computer. Rewrite is optional and, when on, sends only the text to the provider you picked, with your own API key.
macOS, Windows (x64) and Linux (x64). Every feature is on every platform.
Yes on KDE Plasma 6 and GNOME 45+, through the desktop portal. Sway and Hyprland work with XWayland present.
One month of Wispr Flow costs about the same as a lifetime of Wspr. Pay once, own it, no recurring charges, and you get on-device transcription, rewriting and unlimited use.
Built-in dictation has limited accuracy, no custom vocabulary and no AI rewrite. Wspr uses OpenAI's Whisper models across 99 languages, takes custom vocabulary hints, and can rewrite with OpenAI, Anthropic or Google Gemini.
Yes. Turn it on in Settings → General, then hold your hotkey to record and release to transcribe. Quick taps still toggle. Same on macOS, Windows and Linux.
99 languages, detected automatically. Wspr writes in the language you speak.
Free to start on macOS, Windows and Linux.
Linux .deb, checksums and older builds on GitHub