Prepare PDF for AI
Extract PDF content as JSON nodes for RAG / LLM pipelines.
About Prepare PDF for AI
pdf-workshop gets a PDF ready for AI tools in your browser. It takes the text from each page, splits it into chunks of about the size you set (1,000 characters by default) without breaking words, and saves a JSON list where each chunk has an id, its text, the page number and the file name. Nothing is uploaded.
How to use Prepare PDF for AI
- 1Choose or drop the PDF.
- 2Set the Chunk size in characters, or keep 1,000.
- 3Click Prepare for AI and download the JSON file.
Why it runs in your browser
pdf-workshop has no server that receives files. The page loads the code it needs and the work — reading, changing and writing the PDF — happens on your device. Contracts, medical records and ID scans never leave it, there is nothing to sign up for, and it works on networks that block uploads. Large files are limited only by your device's memory.
Questions about Prepare PDF for AI
- Are my files uploaded anywhere?
- No. The work happens in your browser on your own device. pdf-workshop has no server to send files to, so nothing leaves your computer.
- What is in each chunk?
- An id such as node_0, the text, and metadata with page_label (the page number) and file_name. These names match the fields LlamaIndex uses for text nodes, so they are easy to load.
- Can a chunk run across two pages?
- No. Each page is split on its own, so every chunk belongs to exactly one page and its page number is always right.
- Does it work on scanned PDFs?
- Only after OCR. A scan has no text inside, so run OCR PDF first and then prepare the new file.
- Is it free? Do I need an account?
- It is free, with no account, no email address and no trial period. Open the page and use it.
Related tools
Updated September 2026 · How pdf-workshop works · All tools