RagBytes

Turn documents into chunks that are ready to embed.

Upload PDFs, Word documents, web pages, Markdown or plain text. RagBytes strips page clutter, keeps headings, tables and images in context, and returns chunks you can drop straight into a vector store.

How it works

  1. 1.Your documentPDF, Word, web page, Markdown or text
  2. 2.We chunk itBy size, by section, or by meaning
  3. 3.You get chunks backJSON with headings and pages, ready to embed
Ben Merrick
FounderBen MerrickConnect on LinkedIn

7 years of dental school and becoming a doctor, just to find out I want to work in tech. I am open for work.

Documents chunked
–
Chunks made
–

No data is stored

Files are chunked in memory and discarded as soon as the chunks are returned. Privacy

1Choose a chunking tier

See how each tier chunks a document

Watch the sample report being split, one chunk at a time.

Basic fills windows of about 2,000 characters, packing the paragraphs under one heading together and overlapping by 200 characters when it has to split. Every section of this report fits in one window, so it gets one small chunk per heading, and the overview ends up apart from the kit list that belongs with it.

0 of 8 chunks

    2Add your documents

    Drop documents here, or click to browse

    .pdf .docx .html .htm .md .markdown .txt

    Your files are processed in memory and discarded as soon as the chunks are returned. Nothing is stored. Privacy

    No document to hand?Download it