Open this page via a local HTTP server (not file://), otherwise ES modules and the model cache may not work.

Welcome to BrowserLM

Local AI in your browser — chat or ask about your documents (PDF, TXT, CSV). Files stay on device.

Hardware acceleration: Checking...

Choose AI model (LLM): Embedding model:

Download size: ~0 MB

About BrowserLM

BrowserLM is a private, in-browser AI client: chat and RAG with no custom backend. Models run locally in your tab via Transformers.js (ONNX, WebGPU when available, WASM fallback).

Upload TXT, Markdown, CSV, JSON, or PDF and ask questions. Chunks and embeddings are stored in IndexedDB on your device — your files are not sent to a server.

Serve the app over http:// (not file://) so ES modules and the model cache work.

Project on GitHub

Drop files to upload

Text, Markdown, CSV, JSON, PDF

BrowserLM

Local AI

Add documents

Generation (LLM) ...

Search (Embeddings)

Waiting to start...

Loading models...

0 MB / 0 MB (0 MB/s)

0 MB / 0 MB

Remaining: --:-- 0%

AI can make mistakes. Verify facts against your original documents. Project on GitHub