New — Bayut listing scraper is live
Feed your bot.
Skip the chaos.
ChunkIt turns PDFs, websites and real-estate listings into a clean, embedded, queryable vector index — automatically.
12k+
Pipelines run
98.7%
Parse success
42 ms
Avg query
7
Vector stores
How it works
From raw source to vector in seconds
Step 1
Source
PDF, URL, listing
Step 2
Parse
Clean text & tables
Step 3
Chunk
Smart token windows
Step 4
Embed
Multilingual vectors
Step 5
Store
Pinecone, Supabase…
Features
Everything you need, nothing you don't
Real-estate preset
Tuned for Bayut, Dubizzle and Property Finder listings out of the box.
3 sources, 1 click
PDF, URL, or listing scrape — same pipeline, same output schema.
Any bot, any stack
Export JSON or push directly to Pinecone, Qdrant, or Supabase.
Built-in test search
Query your index before you ship. See chunks, scores, metadata.
Multilingual
Arabic + English embeddings work side-by-side without extra config.
Quality metrics
Coverage, duplicate ratio, average chunk size — at a glance.
Ready to feed your bot?
Spin up your first pipeline in under a minute. No credit card required.
Start for free