Very Fast Dictation: a fast speech-to-text tool for MACs

Very Fast Dictation is an open source speech-to-text tool designed for Mac users. It uses fast speech recognition technology to convert what the user says into text in real time, for any scenario where text input is required. Hosted on GitHub and developed by developer Avi Aryan, the project uses Parakeet MLX technology to ensure efficient speech processing. Users only need to press the Control key twice to start or stop recording, making it easy to use and suitable for daily office work, writing or note-taking. The tool supports system-level integration, and text is automatically copied to the clipboard and pasted into the current window to enhance work efficiency. The whole program is lightweight and easy to install, suitable for Mac users who pursue efficiency.

Function List

• Real-time speech to text: Quickly convert speech to text using Parakeet MLX technology.
• Shortcut key operation: Press Control twice to start or stop recording, intuitive operation.
• Clipboard integration: transcribed text is automatically copied to the clipboard and pasted into the active window.
• System-level compatibility: Works in the text box of any Mac application.
• Visual cues: The "Recording..." label is displayed in the center of the screen during recording, providing clear feedback on the status.
• Open source and free of charge: the code is open to the public, allowing users to freely modify and optimize.

Using Help

Installation process

To use Very Fast Dictation, users need to install and configure the environment on their Mac first. Below are the detailed steps:

1. Cloning Codebase
   Open a terminal and run the following command to clone the project locally:

   git clone https://github.com/aviaryan/Very-Fast-Dictation.git
   cd Very-Fast-Dictation
   

2. Installation of dependencies
   Project use uv as a package management tool. First make sure that you have installed the uv(available through pip install uv (Install). Then run the following command to install the dependencies:

   uv sync
   

   This will automatically install a setup that includes parakeet-mlx,pynput cap (a poem) pyside6 The necessary dependency packages are included. Make sure you have a good internet connection, the installation process usually takes only a few minutes.

3. Running the application
   After installing the dependencies, run the following command to start the program:

   uv run main.py
   

   The program will run in the background after launching and the user does not need to open the interface additionally.

Usage

Once installed, Very Fast Dictation is ready to use on your Mac. Here's how to do it:

1. Start recording
   In the text input box of any application (e.g., a text editor, browser, or chat tool), quickly press twice the Control keyRecording..." will appear in the center of the screen to indicate that recording has begun. A "Recording..." prompt will appear in the center of the screen, indicating that recording has begun. At this point, the user can speak directly and the voice will be transcribed into text in real time.
2. Stop recording
   Press again twice as fast Control key, the recording stops. The transcribed text is automatically copied to the clipboard and pasted into the currently active text box. The whole process is smooth without manual copy and paste.
3. status
   When recording, the "Recording..." label in the center of the screen will be displayed continuously to remind the user that the program is working. If the label does not appear, the shortcut key may not be triggered correctly, and it is recommended to check if the program is running in the background.
4. Compatibility Notes
   Very Fast Dictation supports text input scenarios in any Mac application, such as Pages, Word, Notion, or the input box in a browser. Users don't have to switch apps - the tool automatically adapts to the currently focused window.

Featured Function Operation

• shortcut operationThe Control key is designed to be pressed twice to simplify operation, allowing the user to quickly switch between recording states without having to open an additional interface.
• Clipboard integration: Once the transcription is complete, the text is automatically copied to the clipboard and pasted, making it suitable for scenarios where quick notes are needed, such as meeting notes or impromptu ideas.
• real-time transcription: With Parakeet MLX technology, speech-to-text is virtually lag-free, making it ideal for rapid input of long passages of text.
• lightweight design: The program takes up few resources and runs in the background without affecting the performance of other applications.

caveat

• microphone authority: The first time you run it, your Mac system will prompt you to grant microphone access, so make sure you allow it to record properly.
• network requirement: Internet connection is required to install the dependencies, but no network is needed to run it, making it suitable for offline environments.
• Debugging Issues: If the shortcut is not working, check the pynput Depend on whether it is installed correctly, or try restarting the program.
• Custom modifications: As an open source project, users can modify the code as needed, such as adjusting shortcuts or adding new features, as described in the documentation on the GitHub repository.

application scenario

1. proceedings
   In meetings or classes, users can quickly record discussions with Very Fast Dictation. After starting the recording, the tool converts the speech or conversation into text in real time and automatically pastes it into the Notes app, saving time on manual typing.
2. writing aid
   For writers or creators who need to quickly record their inspiration, Very Fast Dictation offers a convenient voice input method. Users simply dictate the content and the text can be entered directly into the writing software, making it ideal for brainstorming or creating a first draft.
3. Accessible input
   For users who are not comfortable typing manually (e.g., those with hand injuries or visual impairments), the tool enhances the accessibility experience by enabling efficient text entry through voice input.
4. multitasking
   When multiple tasks need to be handled at the same time, users can quickly record to-do lists or email content through voice input without having to stop what they're doing and switch to the keyboard.

QA

1. Which operating systems does Very Fast Dictation support?
   It is currently macOS only, designed for Mac users, and is not available for Windows or Linux.
2. Do I need an internet connection to use it?
   Internet connection is required to install the dependency, but no network is needed to run it and offline use is supported.
3. How do I change the default shortcut keys?
   Users need to modify the project code in the pynput Shortcut configuration. This can be found in the GitHub repository's main.py Find the relevant code in the file and adjust it.
4. Does poor recording quality affect transcription?
   Yes, it is recommended to use a clear microphone and record in a quiet environment to ensure transcription accuracy.
5. Can it be used for voice transcription in other languages?
   Currently, the main support is for English transcription, support for other languages depends on the capabilities of the Parakeet MLX model, we recommend checking the official documentation for confirmation.