YASSERRMD

YASSERRMD / barq-web-rag

Public

A fully browser-native RAG application for document Q&A, powered by Rust and WebAssembly with local vector search, embeddings, and in-browser LLM inference.

barq browser-vector-db local-inference rag rag-in-browser
22
6
100% credibility
Found Mar 13, 2026 at 22 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

A browser app for uploading documents like PDFs and Word files to chat with their content using a local AI assistant that runs entirely on your device.

How It Works

1
🔍 Discover Barq RAG

You find this handy web app that lets you upload your own documents and chat with them using a smart helper right in your browser.

2
🚀 Wake up the AI

Click the big button to load the thinking brain – it grabs what it needs once and saves it for future chats so it's quick next time.

3
📤 Add your files

Drag or pick your PDF, Word docs, or text files and drop them in – watch as they're ready to explore in seconds.

4
💬 Ask away

Type any question about your files in the chat box and send it – the answers stream in live like a real conversation.

5
🔍 Check the sources

Tap the source button to see exactly which parts of your files helped make the answer, with match strengths shown.

✅ Get private smarts

Enjoy helpful insights from your docs, all kept safely on your own computer with nothing shared online.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 22 to 22 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is barq-web-rag?

Barq-web-rag is a TypeScript-based, fully browser-native RAG application that lets you upload documents like PDF, DOCX, or TXT and query them via local LLM inference—no servers or APIs needed. It handles everything client-side: document parsing, local embeddings, vector search, and generation with a 1.2B-parameter model over WebGPU. Developers get a ready-to-run chat interface for private, offline document Q&A after a one-time 800MB model download.

Why is it gaining traction?

It stands out by delivering real in-browser LLM inference and local vector search powered by Rust WebAssembly, keeping all data on-device for full privacy in a fully remote setup. No backend hassles mean instant deployment to any static host, with streaming responses and source citations users can expand. The hook? Zero-cost, browser-only RAG that rivals cloud tools without latency or bills.

Who should use this?

Frontend devs building fully kiosk browser apps or github fully remote jobs demos needing offline AI. Indie hackers prototyping document Q&A for fully remote companies, or security pros avoiding data leaks in local inference pipelines. Ideal for quick POCs where fully noded github workflows demand no external dependencies.

Verdict

Try it for bleeding-edge browser RAG experiments—solid docs and Vite setup make local dev easy—but its 22 stars and 1.0% credibility score signal early-stage maturity with experimental WebGPU support. Watch for polish; fork if you need production hardening.

(187 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.