AI-powered voice typing for your entire desktop. Press a hotkey from any app, say what you're thinking, and the text appears at your cursor in seconds. Runs offline with Whisper AI or connects to OpenAI and Gemini for fast, accurate results.
Run everything on your hardware for maximum privacy, or tap into cloud APIs when you need the highest accuracy.
Powered by Whisper AI (via faster-whisper), running entirely on your hardware. Download a model once, then transcribe forever with zero internet.
Send audio to OpenAI Whisper or Google Gemini for top-tier accuracy. Only triggered when you press the hotkey.
A polished interface for managing your settings, history, and transcription preferences.
Idle — sits quietly at the bottom of your screen
Hover — shows your hotkey and how to start dictating
Recording — live audio visualization with cancel and stop
Built for speed and simplicity. No bloat, no setup wizards, no accounts.
Email, Slack, your IDE, Google Docs, a terminal — one global hotkey triggers recording wherever you are. No plugins needed.
Local mode runs Whisper AI entirely on your hardware. No cloud, no network, no exceptions.
Text is transcribed and pasted at your cursor automatically. No copy button, no extra window. Talk, stop, done.
Supports every language Whisper handles, with automatic detection. Switch languages in the tray menu or let it detect.
Every transcription is saved with the original audio recording. Search your history and play back what you said.
Lives in the system tray. A tiny floating pill appears only while recording, with audio bars, cancel, and stop controls.
No account, no sign-up, no configuration required. Just install and go.
Run the installer or extract the portable zip
Ctrl+Shift+R starts recording
Speak naturally — say whatever you're thinking
Text appears at your cursor, already pasted
Anywhere you'd normally type, you can talk instead.
Compose emails, Slack messages, and chat replies at the speed of thought.
Dictate documentation, code comments, and commit messages from your IDE.
Capture ideas into Notion, Obsidian, or any text editor without breaking focus.
Reply in Discord, Teams, or any social app. Long replies are finally easy.
No telemetry, no analytics, no background network calls. We don't collect anything.
No telemetry, crash reports, or usage tracking. Nothing is ever sent to WhisperClick developers or any third party.
Audio is processed entirely on your machine. No internet connection is used. Nothing leaves your computer. Ever.
There is no always-on microphone. Audio is captured only when you explicitly press the hotkey. Nothing else triggers recording.
Free for personal use. Source available. No account required.
Installs to your system with a Start Menu shortcut. Best for most users.
Download .exeUnzip and run. No install needed. Great for USB drives and shared machines.
Download .zipOne executable. Drop it anywhere and double-click to run.
Download .exeFor M1, M2, M3, and M4 Macs (2020 and newer). Mount the DMG and drag to Applications.
Download .dmgFor Intel Macs (2015–2020). Mount the DMG and drag to Applications.
Download .dmgNot sure which Mac you have? Click Apple menu → About This Mac. If it says M1, M2, M3, or M4, choose Apple Silicon.
Portable AppImage for most Linux distributions. Download, make executable, and run.
Download .AppImageYes. Free for personal and non-commercial use. If you use cloud transcription, your API provider may charge a small amount per request, but typical usage is well under $1/month.
In local mode, no. Audio is processed entirely on your hardware. In API mode, audio is sent to your chosen provider only when you press the hotkey. Nothing is stored after transcription returns.
Yes. Download a Whisper model once, then all transcription happens locally with zero internet. Models range from "tiny" (fast, lower accuracy) to "large-v3" (slower, highest accuracy).
Encrypted at rest using your OS native credential store (Windows Credential Locker, macOS Keychain, or Linux secret service). Never stored in plain text.
Some apps with elevated privileges may block simulated keystrokes. Switch to "Copy only" mode in Settings and paste manually with Ctrl+V (Cmd+V on Mac).
macOS Gatekeeper may block unnotarized apps. Right-click the app, choose "Open", then click "Open" in the dialog. You only need to do this once.
WhisperClick is source-available under CC BY-NC-SA 4.0. Clone the repo, build from source, or read every line of code that touches your microphone.
View on GitHub# Clone and run from source git clone https://github.com/Zbrooklyn/WhisperClick-Desktop-App.git cd WhisperClick-Desktop-App python -m venv venv venv\Scripts\activate pip install -r requirements.txt python src/main.py
Everything you need to get up and running.
Set up your API key, choose a model, and start transcribing in under two minutes.
Read the guideFound something that doesn't work right? Open an issue and we'll take a look.
Open an issueBrowse the codebase, submit pull requests, or build from source yourself.
View on GitHub