PressScribe

A desktop application for Windows (Mac and Linux - use Python script or build your own distributive) that allows you to transcribe your speech in real-time, polish the text using AI, and manage your transcriptions with a simple and intuitive interface.

==Free API key from Google AI GEMINI in 2025:== How To Get Your FREE Google Gemini API Key (2025)

How to use - see the screenshot:

How to get key (I still highly recommend to watch that youtube video above):

• Login in https://aistudio.google.com/
• Click GET API Key in the top right corner
• Go through the steps (it is easy) copy the key and paste in the app.
• How much will you pay - unless for any reason you don't get FREE API KEY from Google GEMINI, you it is still cheap as dirt.
• TRANSCRIPTION IS ALWAYS FOR FREE! You don't need a key for it, you need key only for Polishing the text.
• You can install LM Studio or other local AI programs and if you machine is powerful enough you don't need API key at all.

Tech stuff (don't read it you are not a nerd)

• Interactive Transcription: Use your microphone to transcribe speech directly into the editor at the cursor's position. Select text to replace it with a new transcription.
• Open Audio: Import an audio file (WAV, MP3, M4A, and more) and transcribe it with your selected transcription service.
• AI-Powered Polishing: Polish the entire text or just a selection using an AI (Gemini or a local model) to correct grammar, improve phrasing, and fix typos.
• Polish + Translate: Proofread and translate raw text into a chosen language with Gemini (Settings → Choose Translate Language).
• Saved Notes: Auto-save polished text into a notes library (on by default). Switch to Saved Notes to open, edit, copy, or delete notes. Manual Save buttons work on both raw and polished panels.
• Modern Theming: Choose between beautiful, consistent light and dark themes.
• Flexible AI Options: Easily switch between Google's Gemini API and a local AI model running on your machine (e.g., via LM Studio).
• Multiple Transcription Engines: Choose between Google Speech Recognition, local Faster-Whisper, or a Qwen 3 ASR server running on your LAN.
• Session Management:
  • Save & New: Save your current transcription and the polished text to a JSON file and clear the editors for a new session.
  • Open: Load a previously saved session to continue your work.
• Persistent Settings: Your API key, AI service preference, theme choice, translate language, and autosave preference are saved automatically between sessions.

Setup and Installation

Follow these steps to run the application from the source code.

1. Prerequisites

• It is tested on Python 3.13.3
• A working microphone.

3. Install Dependencies

Open a command prompt or PowerShell terminal, navigate to your project folder, and run the following command to install the required Python libraries:

pip install -r requirements.txt

4. Running the Application

To run the application directly from the script, use the following command in your terminal:

python transcriber.py

5. Qwen 3 ASR Server Setup

• The app now supports a third transcription backend: Qwen 3 ASR Server.
• Configure it in Settings > Transcription Service > Qwen 3 ASR Server.
• Set the server address in Settings > Set Qwen ASR Server URL....
• The default server on this machine is http://192.168.1.114:8711/transcribe.
• This server is designed to be reached through the machine's 192.168.x.x LAN address, not through 127.0.0.1.
• The server deployment in this environment lives under /mnt/merged_ssd/vocolibri_production/transcription.

Creating a Standalone Executable (.exe)

You can package the application into a single .exe file that can be run on any Windows computer, even without Python installed.

1. Install PyInstaller:
   If you haven't already, install PyInstaller:
   pip install pyinstaller

2. Build the Executable:
   In your terminal, from the project directory, run the following command. It is recommended to run PyInstaller as a Python module to avoid potential path issues.
   python -m PyInstaller --onefile --windowed --name="PressScribe" --icon="icon.ico" transcriber.py
   or (if you want to create with embedded icon image):
   python -m PyInstaller "PressScribe.spec"

   • --onefile: Bundles everything into a single .exe file.
   • --windowed: Prevents a console window from appearing when the application runs.
   • --name: Sets the name of the final executable.

3. Find Your Application:
   Once the process is complete, you will find PressScribe.exe inside a new folder named dist. You can share this file with others.