🎤 Gemini Dictate

A Chrome extension that enables speech-to-text transcription using Google's Gemini AI. Activate with a keyboard shortcut and dictate directly into any text input field on the web.

Features

• ⌨️ Keyboard Shortcut Activation: Press Ctrl+Shift+S (or Cmd+Shift+S on Mac) to start dictating
• 🎯 Smart Input Detection: Only works when a text input field is focused
• 🔄 Real-time Transcription: Powered by Gemini 2.0 Flash for accurate speech recognition
• 👁️ Visual Feedback: Clear indicators when dictation is active
• 🛡️ Privacy-Focused: Your API key stays local, audio sent directly to Gemini
• 🎨 Beautiful UI: Modern, gradient-styled options page

🗂️ Project Versions

This repository contains two versions of Gemini Dictate:

1. Chrome Extension: Works directly in your browser. Supports built-in Chrome Speech-to-Text (No API key required) or Gemini AI (Premium accuracy).
2. macOS Electron App: A system-wide application that works in any Mac app (Notes, Slack, etc.). Requires a Gemini API key.

---

🔑 1. Setup Your Gemini API Key (Optional for Extension)

You can skip this part if you only want to use the built-in Chrome Speech-to-Text engine in the Chrome Extension. You only need an API key if you want to use the high-accuracy Gemini engine in the Chrome Extension, or if you are using the Electron App.

1. Create a Google Cloud Project: Follow the Google Cloud documentation to create a new project.
2. Get your API Key:
   • Recommended: Visit Google AI Studio to generate a free API key.
   • Alternative: Use the Google Cloud Vertex AI documentation for enterprise-grade setup.
3. Copy the key for the next steps.

---

🌐 2. Chrome Extension

The Chrome extension can use your browser's built-in transcription engine immediately, or you can supply a Gemini key for significantly better results.

Installation

1. Download or clone this repository.
2. Open Chrome and navigate to chrome://extensions/.
3. Enable "Developer mode" (toggle in the top-right corner).
4. Click "Load unpacked".
5. Select the root folder of this repository (where manifest.json is located).
6. (Optional) Click the extension icon in Chrome's toolbar, select Options, enter your Gemini API key, and select the Gemini engine.

Usage

1. Focus: Click into any text input field on a webpage.
2. Activate: Press Cmd+Shift+S (Mac) or Ctrl+Shift+S (Windows).
3. Dictate: Speak clearly. A visual indicator will appear.
4. Stop: Press the shortcut again or click away.

---

💻 3. macOS Electron App

A native application for system-wide dictation. Note: This version requires a Gemini API key.

Installation

1. Ensure you have Node.js installed.
2. Open your terminal and navigate to the electron-app directory:

   cd electron-app

3. Install the dependencies:

   npm install

Usage

1. Start the app:

   npm run start

2. Configure: On first launch, enter your Gemini API key in the Settings window.
3. Key Bindings: Use Command+Shift+S (default) while focused on any application.

Building & Packaging

To create a standalone .app bundle:

npm run make

---

🛠️ Development & Troubleshooting

Project Structure

• /: Root directory contains the Chrome Extension files.
• /electron-app: Contains the native macOS application.

Troubleshooting

• Microphone Access: Ensure Chrome or the Electron app has permission to access your microphone in System Settings.
• Engine Selection: If you haven't provided an API key in the extension settings, it will fallback to the default Chrome Web Speech engine.
• API Errors: Use the "Test Connection" button in the Settings page to verify your key if using Gemini/Chirp.

📄 License

MIT License - feel free to use, modify, and distribute!

---

Made with ❤️ using Gemini AI