Think It. Say It. Paste It.

The privacy-first utility that empowers your workflow.
Professional Offline Transcription. High-Fidelity AI Speech.

System Requirements macOS 13.0+ (Apple Silicon) ~4GB Disk Space for AI Models v3.4.5

* OneSentence uses AI models which can occasionally make mistakes.

Private by Design

Your voice and text never leave your device. OneSentence runs entirely offline using local binaries. No cloud. No tracking.

Neural Engine Native

Optimized for Apple Silicon (M1-M4). High-performance AI utilizes the Neural Engine and GPU for zero-latency processing without draining your battery.

Device Precision

Switch between microphones and speakers directly from the menu bar. Perfect for power users with complex audio setups.

Zero Subscriptions

Tired of monthly fees? OneSentence is a comprehensive utility with a simple model. One-time purchase, no subscriptions.

Universal Voice Input

Professional Voice Input for any application.

OneSentence provides a seamless interface for your favorite tools. Dictate directly into ChatGPT, Claude, Gemini, or any professional editor with unprecedented accuracy. It is the most natural and efficient way to interact with your digital workspace.

"こんにちは、契約書を送ってください。"
Whisper (Local) Japanese Mode
こんにちは、契約書を送ってください。
Powered by Whisper Large-v2

Professional dictation.
Completely offline.

Transform your speech into precise text instantly. Powered by OpenAI's Whisper Large-v2 model, OneSentence delivers world-class transcription accuracy without ever needing an internet connection.

  • Model Versatility Choose between Tiny, Base, Small, Medium, and Large-v2 models to balance speed and accuracy.
  • Non-Destructive Workflow Our Clipboard Snapshot technology preserves your history, instantly restoring your previous clipboard after transcription.
  • Polyglot Accuracy Automatic language detection and transcription for dozens of global languages.
Supertonic Engine

Stop scanning.
Start listening.

Transform long-form content into a professional narrated experience. OneSentence uses advanced Supertonic diffusion models to provide a pro-audio listening experience that sounds like a studio narrator, not a computer.

Pro Audio Control

Fine-tune the "Diffusion Steps" to choose between ultra-fast generation or studio-grade quality.

System-Wide

Works in any application with a simple global hotkey.

Supertonic models are trained only on English data; outputs in other languages are unsupported and may be unintelligible or offensive.

Reading (English US)

Reading Selection

Using "Voice M1" (Supertonic) at 1.2x speed.

We don't want your data.

Your thoughts belong to you. OneSentence is built so we literally cannot see your data. The application is 100% offline. No cloud. No tracking. No compromise.

Metric
OneSentence Policy
In-App Data Collection
None
In-App Cloud Uploads
None (100% Offline)
Analytics & Tracking
None
Optional Web Feedback
Browser-based only

Frequently Asked Questions

Why are the AI models downloaded separately?

To maintain a lightweight app package, the high-performance AI models are downloaded securely on-demand during initial setup. OneSentence requires approximately 4GB of disk space to house these professional-grade local engines (Whisper and Supertonic).

Does this work completely offline?

Yes. Once the models are downloaded, you can be in the middle of a desert with no Wi-Fi and OneSentence will work perfectly. Your voice and text data never leave your machine.

How do I access the interface?

OneSentence lives in your menu bar (top right of your screen). Click the waveform icon to access settings, or use the global hotkeys (Command+5 and Command+6 by default) to trigger functions instantly.

How do I manage disk space or uninstall?

OneSentence includes built-in maintenance tools. In Settings > Maintenance, you can "Prune Models" to free up space while keeping your settings, "Reset Settings" to start fresh, or perform a "Full Uninstall" which safely removes all application data. Both uninstall options offer an optional web-based survey to securely send us your feedback.