Features
Voice-to-text built for macOS developers
System-level dictation with push-to-talk control, OpenAI Whisper accuracy, and direct text injection into any application.
Global Push-to-Talk Shortcut
Configurable system-level key binding that works in any application. Hold the shortcut to record, release to transcribe. No window switching required.
How it works
VibeWhisper registers a global keyboard shortcut at the system level using CGEvent taps. When you hold the configured key, audio capture begins immediately via AVAudioEngine. Release the key and the audio buffer is sent to the OpenAI Whisper API. The shortcut works in any application — your IDE, browser, terminal, chat apps, or email client.
How it works
VibeWhisper registers a global push-to-talk shortcut using macOS CGEvent taps. Hold the key to start recording, release to send audio to the OpenAI Whisper API for speech to text transcription. The shortcut is system-wide and works in any application.
Use case
You're writing code in your IDE and need to dictate a commit message. Hold your shortcut, speak the message naturally, release — the transcribed text appears at your cursor without leaving your editor.
OpenAI Whisper
Low-latency, accurate speech-to-text powered by OpenAI's Whisper API. Supports 99+ languages out of the box.
How it works
Audio is sent directly to the OpenAI Whisper API using your own API key. Whisper is a general-purpose speech recognition model trained on 680,000 hours of multilingual data. It supports transcription in 99+ languages with high accuracy on technical terms, code-related vocabulary, and natural speech patterns.
How it works
Audio is sent directly to the OpenAI Whisper API using your own API key. Whisper is a state-of-the-art voice recognition model trained on 680,000+ hours of multilingual data, delivering accurate speech to text in 99+ languages — including technical vocabulary and code-related terms.
Use case
You're dictating a detailed prompt to an AI coding assistant. Whisper accurately transcribes technical terms like "API endpoint," "JWT token," and "PostgreSQL migration" that generic dictation software would miss.
Accessibility API Text Injection
Text inserted directly into the focused text field via macOS Accessibility APIs. No clipboard hijacking, no paste shortcuts.
How it works
After transcription, VibeWhisper uses the macOS Accessibility API (the same system used by assistive technologies) to inject text directly at the cursor position in the currently focused text field. This means no clipboard hijacking — your clipboard contents remain untouched.
How it works
After transcription, VibeWhisper injects text directly into the focused text field using the macOS Accessibility API. This is the same system-level API used by assistive technologies. Your clipboard contents remain untouched — no paste workarounds required.
Use case
You're composing a Slack message while a code snippet sits on your clipboard. VibeWhisper inserts the dictated text without overwriting your clipboard, so you can still paste your code afterward.
Secure Keychain Storage
Your OpenAI API key is stored in macOS Keychain. It never leaves your machine and is never transmitted to VibeWhisper servers.
How it works
Your OpenAI API key is stored in the macOS Keychain, the same secure storage system used for passwords, certificates, and system credentials. The key never leaves your machine and is never transmitted to any VibeWhisper server. Only direct API calls to OpenAI use the key.
How it works
Your OpenAI API key is stored in the macOS Keychain, the same encrypted storage used for passwords and certificates. The key never leaves your machine and is never sent to VibeWhisper servers — only direct API calls to OpenAI use it.
Use case
You enter your API key once during setup. From that point on, it stays encrypted in the macOS Keychain. Even if you share your Mac with others, your key remains protected by the OS-level security layer.
Menu Bar App
Lightweight menu bar presence. Always available, zero overhead. Configure your shortcut and API key, then forget about it.
How it works
VibeWhisper runs as a lightweight menu bar application with no Dock icon. It uses minimal system resources and stays out of your way. Configure your shortcut key and API key once, and the app is ready whenever you need it.
How it works
VibeWhisper runs as a lightweight menu bar application with no Dock icon. It uses minimal system resources and stays out of your way. Configure your shortcut key and API key once, and the voice typing app is ready whenever you need it.
Use case
You start your Mac in the morning and VibeWhisper is already running in the menu bar. No windows to open, no apps to launch. Just hold your shortcut anywhere on the system and start dictating.
Integrations
Works with every app on macOS
VibeWhisper is a system-level speech to text app that uses the macOS Accessibility API for universal compatibility. If you can type in it, you can dictate into it.
Comparison
How VibeWhisper compares to other dictation software
| Feature | VibeWhisper | macOS Dictation | Subscription Apps |
|---|---|---|---|
| Push-to-Talk | Global system-level shortcut via CGEvent taps | Limited — keyboard toggle or voice-activated | Varies by product |
| Voice Recognition Accuracy | OpenAI Whisper — trained on 680,000+ hours of multilingual data | Apple on-device model — struggles with technical terms | Proprietary models — accuracy varies |
| Cost | $19 one-time + ~$0.006/min API usage | Free (built-in) | $10–20/month subscription |
| Privacy | Audio sent directly to OpenAI with your key — no intermediary | Processed on Apple servers | Routed through vendor servers |
| Languages | 99+ languages via OpenAI Whisper | 60+ languages | Varies — typically 20-100 |
FAQ
Frequently asked questions
How does the push-to-talk shortcut work?
VibeWhisper registers a global keyboard shortcut via macOS CGEvent taps. Hold the key to record, release to transcribe. The shortcut works system-wide in any application — no need to focus VibeWhisper first.
How accurate is the voice recognition?
VibeWhisper uses OpenAI Whisper, trained on 680,000+ hours of multilingual audio. It handles technical vocabulary, accents, and natural speech patterns with high accuracy across 99+ languages.
What does VibeWhisper cost?
VibeWhisper is a $19 one-time purchase — no subscription. You bring your own OpenAI API key and pay OpenAI directly at approximately $0.006 per minute of audio. There are no hidden fees or markup.
Is my voice data private?
Yes. Audio is sent directly to the OpenAI Whisper API using your own key. VibeWhisper never stores, logs, or processes your audio. Your API key stays in the macOS Keychain and is never transmitted to our servers.
Does it work in my code editor or IDE?
VibeWhisper works in any text field accessible via the macOS Accessibility API. This includes VS Code, Cursor, Xcode, IntelliJ, terminals, browsers, and virtually every other macOS application.
Start dictating in any app on macOS
One-time purchase. No subscription. No account required. Just a shortcut key and your voice.