Weby is a revolutionary browser extension that brings the power of Large Language Models (LLMs) directly into your browser—100% locally. No API keys, no subscriptions, and total privacy. Weby indexes the pages you visit in real-time, allowing you to chat with your web content, summarize articles, and extract insights without your data ever leaving your machine.
- 🔒 100% Local Privacy: Everything runs on your hardware. No data is sent to external servers or AI providers.
- 🧠 Intelligent Web RAG: High-performance Retrieval-Augmented Generation (RAG) that understands the specific context of the website you are currently viewing.
- 🔄 Tab-Synced Conversations: Switch tabs, and Weby automatically brings up the last conversation you had for that specific URL.
- ⚡ Supercharged Sidebar: A sleek, ChatGPT-like interface that lives in your browser's side panel for instant access.
- 🌐 Web Research Mode: Need more info? Weby can search the web locally to supplement its knowledge.
- 🛠️ MCP Integration: Connect to Model Context Protocol (MCP) servers like Notion, Google Drive, or local tools to extend your AI's capabilities.
- 🎙️ Voice Mode: Talk to your web content and hear responses back with built-in Speech-to-Text and Text-to-Speech.
Since Weby runs high-performance WASM and WebGPU models locally, it is currently distributed as a developer-mode extension for maximum performance.
- Download the Latest Build: Go to the Releases page and download
weby-extension.zip. - Unzip the File: Extract the contents to a folder on your computer.
- Open Extensions Page: In Chrome (or any Chromium browser), go to
chrome://extensions. - Enable Developer Mode: Toggle the switch in the top right corner.
- Load Unpacked: Click the "Load unpacked" button and select the folder you just extracted (the one containing
manifest.json). - Start Chatting: Open the side panel by clicking the Weby icon and start chatting!
Weby uses a sophisticated local pipeline to provide high-quality AI responses:
- Extraction: It uses
Turndownto convert the active tab's HTML into clean Markdown. - Indexing: A local embedding model (via
WllamaorWebLLM) chunks and vectors the text into an IndexedDB. - Retrieval: When you ask a question, Weby performs a hybrid semantic search to find the most relevant parts of the page.
- Generation: A local LLM (like Qwen 3 0.5B or - Llama 3.2 1B) processes the context and generates an answer right in your browser.
- Frontend: Preact + Tailwind CSS
- Local AI Engines: WebLLM (WebGPU) & Wllama (WASM)
- Database: IndexedDB (via
idb) - Communication: Chrome Extension Bridge
- Workflow: Automated CI/CD via GitHub Actions
We welcome contributions! Whether it's reporting a bug, suggesting a feature, or submitting a pull request, your help makes Weby better for everyone.
- Fork the Project
- Create your Feature Branch (
git checkout -b feature/AmazingFeature) - Commit your Changes (
git commit -m 'Add some AmazingFeature') - Push to the Branch (
git push origin feature/AmazingFeature) - Open a Pull Request
Distributed under the GNU GPLv3.0 License. See LICENSE for more information.
Built with ❤️ by Inled Group