Turn speech into text using fast, private, on-device recognition. Reduce typing strain with hands-free text entry in 30+ languages. No subscription. No cloud.
Everything you need to convert speech into text comfortably and privately.
Speech recognition runs locally on your Mac. Your voice never leaves your device. No cloud, no servers, no data collection.
Speak naturally and convert speech to text across languages. Auto-detect or choose your preferred language.
You can improve grammar or clarity after transcription if desired. Uses your own AI provider—completely optional.
Start voice input easily when speech is more comfortable than typing. Hold to speak or click to toggle.
Clear visual and sound cues when voice input starts or stops. Always know when the app is listening.
No internet connection required for transcription. Download once, use anywhere—on a plane, in a café, or offline.
Voice to text has never been this effortless.
Press ⌥ Option + Space or use the floating pill
Talk naturally in any language
Speech is entered as text where you're typing
Speech recognition runs locally on your Mac. No cloud processing required.
Designed for privacy-first voice input.
No subscriptions. No hidden fees. Pay once, use forever.
Everything you need to type with your voice
One-time purchase • Free updates forever
Most voice apps charge $15/month per user. Do the math.
With a family of 4? Subscriptions cost $3,600 over 5 years. VoiceLah! stays $29.99.
Want smart grammar fixes and translation to any language? Bring your own API key. You control the cost, you own the data.
Yes! VoiceLah! uses Whisper AI that runs entirely on your Mac. Once downloaded, you can use it anywhere without internet. Only the optional AI polish feature requires a connection.
VoiceLah! requires macOS 14 (Sonoma) or later. It works best on Apple Silicon Macs (M1, M2, M3, M4) with near-instant transcription. Intel Macs are supported but will be slower.
Absolutely. All speech recognition happens locally on your Mac using Whisper AI. Your voice never leaves your device. We don't have servers, don't collect data, and don't even require an account.
AI polish is optional and uses AI providers (OpenAI, Anthropic, etc.) to improve your text—fixing grammar, punctuation, and clarity. You bring your own API key, so you control the cost. It also enables translation to any language beyond English.
We believe in fair pricing. Since VoiceLah! runs locally and doesn't require our servers, there's no reason to charge you monthly. Pay once, use forever, get free updates.
VoiceLah! uses OpenAI's Whisper model, one of the most accurate speech recognition systems available. You can choose different model sizes—larger models are more accurate but slower. Most users find the "Balanced" setting perfect.
1. Go to platform.openai.com and create an account
2. Add a payment method in Billing
3. Go to API Keys and create a new key
4. Copy the key into VoiceLah! settings
Cost: GPT-5o Nano costs ~$0.10 per million input tokens. For typical voice corrections, expect less than $0.01 per day of heavy use.
1. Go to console.anthropic.com and create an account
2. Add a payment method in Settings → Billing
3. Go to API Keys and generate a new key
4. Copy the key into VoiceLah! settings
Cost: Claude Haiku costs ~$0.25 per million input tokens. For text corrections, expect less than $0.01 per day even with heavy use.
1. Go to openrouter.ai and sign up
2. Add credits in your account (starting from $5)
3. Go to Keys and create a new API key
4. Copy the key into VoiceLah! settings
Why OpenRouter? Access to 100+ models from one API. Use Claude, GPT, Llama, Mistral and more. Great for trying different models or finding the best price/performance ratio.
1. Download and install Ollama for Mac
2. Open Terminal and run: ollama pull llama3.2
3. In VoiceLah! settings, select "Ollama" as your AI provider
4. That's it! No API key needed.
Cost: FREE forever. Ollama runs AI models locally on your Mac. Recommended models: llama3.2 (fast), mistral (balanced), or gemma2 (lightweight). Requires ~4-8GB RAM depending on model.
Join thousands of Mac users who use voice input every day.
Download on the App Store