🧠 Mistral Chrome Extension

A lightweight Chrome extension that lets you highlight text on any webpage, send it to a locally running Mistral model (via Ollama), and view the response as a tooltip — all without the cloud. Fast, private, and offline.

---

🛠️ Prerequisites

This project assumes you're using Windows, macOS, or Linux with basic Python and Git knowledge.

✅ Tools You Need:

• Ollama (to run Mistral locally)
• Python 3.8+ (to run the local API server)
• Chrome (or Edge, Brave, etc.)
• Git

---

🚀 Step 1: Install Ollama and Mistral Model

1. Install Ollama

Visit: https://ollama.com/download

Or use the command below on macOS/Linux:

curl -fsSL https://ollama.com/install.sh | sh

On Windows, just run the .exe installer.

2. Pull the Mistral Model

Open a terminal and run:

ollama run mistral

This will:

• Download the Mistral 7B model (~4GB)
• Launch an interactive local chat
• Start a local inference server at http://localhost:11434

Leave this terminal running.

🧠 Step 2: Run the Local Mistral API Server (Flask)

1. Clone this repo

git clone https://github.com/kcho2027/Mistral-Extension.git
cd Mistral-Extension

2. Install Python dependencies

pip install flask requests

3. Run the local server

python mistral_server.py

This exposes a helper API at:

http://localhost:11435/ask

Leave this terminal running too.

🌐 Step 3: Load the Chrome Extension

1. Open Chrome and go to:

chrome://extensions

2. Enable Developer Mode (top right)

3. Click "Load Unpacked"

Select the Mistral-Extension folder containing:

• manifest.json
• background.js
• icon.png

You should now see the extension installed.

🧪 How to Use

1. Open any webpage.
2. Highlight some text.
3. Right-click → select "Ask Mistral about this text".
4. The tooltip will appear in the top-right corner with Mistral's response.

🖱️ The response box supports scrolling and auto-disappears after 15 seconds.

📂 Folder Structure

Mistral-Extension/
├── manifest.json    # Chrome extension config
├── background.js    # Extension logic & tooltip injection
├── icon.png         # Extension icon
├── mistral_server.py # Flask API server that connects to Ollama
└── README.md        # You're reading this

✅ Features

• ✅ Completely local and offline
• ✅ Uses the powerful Mistral 7B model
• ✅ Runs via Ollama
• ✅ No external API keys or OpenAI account needed
• ✅ Auto-injected scrollable response box
• ✅ Supports any Chromium browser

🔒 Privacy

All your text, prompts, and model responses stay on your device — nothing is sent to any cloud service. Perfect for private research and local AI tooling.

🧩 Coming Soon Ideas

• Show response near cursor instead of top-right
• Keep chat history per session
• Add keyboard shortcut (e.g. Ctrl+M) to trigger prompt
• Add a dismiss (X) button to tooltip
• Automatically resize response box for long outputs
• Add multi-model support via Ollama (mistral, llama2, etc.)

🙌 Author

Created by Kyoungbin Cho Pull requests, ideas, and contributions welcome!

📜 License

MIT License. Free for personal or commercial use.

---

Let me know if you want help formatting this as a GitHub Pages page, generating preview badges, or adding images!