AI Chat with your Website

hub

Crawl an entire website — docs, blogs, wikis — and chat with AI to ask questions across all pages. Get cross-page summaries, find specific information, and extract insights.

shield Auto-deleted after session
token Loading quota...

hub Website

Try:
link
30

Only crawl pages under the URL path you enter

Skip pages the site asks crawlers not to visit

Paste a URL to start crawling. We'll extract text from all linked pages within the same site.

forum Chat

chat_bubble

Let's chat about a website

Enter a URL in the left panel to crawl a website and start chatting across all its pages.

Crawl a website to start chatting

How It Works

travel_explore

1. Enter a URL & Crawl

Paste any website URL and click Crawl. We'll follow links breadth-first, extracting text from each page. Watch real-time progress as pages are discovered.

memory

2. Smart Indexing

Text from each page is chunked, deduplicated across pages (no repeated navs/footers), and embedded into a vector index. Each chunk remembers which page it came from.

chat

3. Chat with Source Citations

Ask questions and the AI finds the most relevant chunks across all pages. Answers include which page the information came from — full source attribution.

Use Cases

📚

Documentation Sites

Crawl your framework's docs and ask "How do I set up authentication?" — the AI searches across all pages to find the answer with source links.

📝

Blog Archives

Index an entire blog and ask questions that span multiple posts. "What are the author's main arguments about testing?" pulls from across all articles.

🏢

Company Knowledge Bases

Crawl internal wikis or knowledge bases and find answers across hundreds of articles. Perfect for onboarding or quick reference.

🔬

Research & Tutorials

Index multi-page tutorials or course sites. Ask targeted questions and get answers with references to the specific lesson page.

Try an Example

Click any card below to start crawling a website and chat with its content.

code

Python Tutorial

The official Python tutorial covering data structures, modules, I/O, and more. Ask about classes, list comprehensions, or error handling.

language docs.python.org
travel_explore Click to crawl →
school

MDN: JavaScript First Steps

MDN's beginner JavaScript tutorial module. Ask about variables, strings, arrays, or troubleshooting.

language developer.mozilla.org
travel_explore Click to crawl →
article

Engineering Blog

A small engineering blog. Ask questions that span multiple posts, or get a summary of all topics covered.

language intervu.dev
travel_explore Click to crawl →

Frequently Asked Questions

How does this differ from Webpage Chat? expand_more
Webpage Chat loads and analyzes a single page. Website Chat crawls multiple pages from the same site (up to 100), indexes all the text into a searchable vector database, and answers questions by searching across all pages — with source citations telling you which page each answer came from.
Is my data safe? expand_more
Yes. We crawl public pages, extract text, and store it temporarily in server memory for your chat session. Sessions are automatically deleted after 30 minutes of inactivity or 2 hours maximum. We never persist your data to disk.
Does it respect robots.txt? expand_more
Yes, by default. The crawler checks the site's robots.txt file and skips pages that are disallowed. You can toggle this off if you need to crawl pages blocked by robots.txt for private analysis (e.g., your own site's docs behind a robots.txt rule).
How many pages can I crawl and chat messages can I send? expand_more
You can crawl up to 100 pages per crawl and perform 3 crawls per day. Chat messages are limited to 20 per day, shared across all AI chat tools. Both quotas reset at midnight.