Free Chunk Testing Tool
Enter any URL to see how different AI systems chunk your content for RAG retrieval. Compare 9 strategies and 4 AI presets. Completely free. No sign-up required.
What is content chunking?
Before an AI system can retrieve and cite your content, it needs to break it into searchable pieces. The way it chunks your content directly affects whether it gets cited or ignored.
Page content is fetched
The page is scraped and converted to clean markdown — stripping navigation, ads, and boilerplate to leave only the useful content.
Content is split into chunks
The markdown is divided into smaller pieces using a chunking strategy. Different strategies produce different chunks — some respect headings, others use fixed token windows, and hybrid approaches combine both.
Chunks become searchable
Each chunk is embedded as a vector. When a user asks a question, the AI searches these vectors to find the most relevant chunks. Badly sized or contextless chunks get poor match scores and are never retrieved.
The chunking strategy can make or break your content’s retrievability. This simulator lets you see exactly how each approach treats your pages.
Go deeper with Content Lab
Content Lab adds semantic scoring, entity extraction, and passage highlighting — so you can see exactly which chunks an AI would retrieve for any query.
Learn more View plans