Effortless Voice-to-Text for macOS, Powered by On-Device Whisper AI.
English | 简体中文
SayIt is a lightweight, privacy-focused macOS menu bar application that transcribes your voice into text instantly. By leveraging OpenAI's Whisper model running locally via whisper.cpp, SayIt ensures your data never leaves your machine while providing lightning-fast, high-accuracy transcriptions.
- Global Hotkey: Start and stop recording from any application with a single key combination (Default:
⌥ Space). - Privacy First: 100% on-device processing. No cloud APIs, no data collection.
- Seamless Workflow: Captured text is automatically copied to your clipboard, ready to paste.
- Smart HUD: A non-intrusive floating HUD shows transcription progress in real-time.
- Elegant UI: Native macOS look and feel with a refined popover for settings and status.
- Whisper Integration: Supports multiple Whisper models (Small, Base, Tiny) for optimal performance.
- Intelligent Audio Routing: Optimized for "Clamshell Mode". Automatically switches to your iPhone microphone via Continuity when the Mac lid is closed and the built-in mic is disabled.
- macOS 15.0 or later.
- Xcode 16.0 or later.
-
Clone the repository:
git clone https://github.com/princeniu/SayIt.git cd SayIt -
Initialize Submodules: SayIt uses
whisper.cppas a submodule.git submodule update --init --recursive
-
Open in Xcode:
open SayIt/SayIt.xcodeproj
-
Build and Run: Select the SayIt scheme and press
⌘R.
- Grant Permissions: On first launch, SayIt will request Microphone and Speech Recognition permissions.
- Download Model: If using Whisper for the first time, download the recommended model via the popover.
- Record: Press
⌥ Spaceto start recording. A level indicator will appear in the menu bar. - Stop: Press
⌥ Spaceagain. Transcription starts automatically. - Paste: Once the HUD shows "Copied!", your text is in the clipboard. Just
⌘Vwhere you need it.
SayIt is built with modern Apple technologies:
- SwiftUI: For a fluid, reactive user interface.
- Combine: For robust state management and asynchronous data flow.
- whisper.cpp: High-performance C++ implementation of OpenAI's Whisper.
- CoreAudio: For low-latency audio capture.
Contributions are welcome! Please feel free to submit a Pull Request.
- Fork the Project.
- Create your Feature Branch (
git checkout -b feature/AmazingFeature). - Commit your Changes (
git commit -m 'Add some AmazingFeature'). - Push to the Branch (
git push origin feature/AmazingFeature). - Open a Pull Request.
xcodebuild -project SayIt/SayIt.xcodeproj -scheme SayIt testDistributed under the MIT License. See LICENSE for more information.
Made with ❤️ for the macOS Community.
