RAG Chatbot Demo

Upload a document and ask questions about it

LLM: initializing

Embeddings: loading

No document

No documents loaded

Response Mode:

Strict

+ General Knowledge

Upload a document to see embeddings

Settings

LLM Provider

Local Model (WebLLM) WebGPU

Runs Llama 3.2 3B locally in your browser. First load downloads ~1.8GB model (cached for future use).

Model cached

Anthropic API (Claude)

Use Claude via Anthropic's API. Requires your own API key.

Your key is stored locally and never sent to our servers.

Browser Requirements

Welcome to RAG Chatbot

This demo lets you upload documents and ask questions using AI. Choose how you'd like to power the AI:

Download Local Model WebGPU

Download Llama 3.2 3B (~1.8GB) to run entirely in your browser. Private & offline-capable after download.

WebGPU not available in this browser

Use Anthropic API (Claude)

Use Claude instantly with your API key. No download required.

Stored locally, never sent to our servers.

Chat History

No saved chats yet

Document Preview

This is what the LLM will see:

Text content ready for processing.

Downloading Model

Llama 3.2 3B Instruct (~1.8GB)

0%

Initializing...

The model will be cached for instant loading on future visits.