How Do You Prepare a PDF for ChatGPT, Claude or RAG?
You prepare a PDF for an AI tool by converting it with QuickCatPDF's PDF to Markdown for LLMs and clicking Convert to Markdown. You get clean Markdown with headings, lists and tables intact, a short header naming the source, and a comment marking where each page starts, plus an estimate of how many tokens the text uses.
The PDF is converted in your browser, so a confidential document isn't uploaded anywhere before you decide where to paste it. It's free, with no sign-up.
How to Convert a PDF for an AI Tool With QuickCatPDF
Preparing a PDF takes four steps.
- Click Select PDF files or drop your PDFs on the page (up to 20).
- Keep Front matter and Page markers as comments ticked so the AI tool knows the source and pages.
- Type a range in Pages if you only need part of the PDF.
- Click Convert to Markdown, then download the file or copy it from the preview.
Why Convert a PDF to Markdown Before Using AI?
Language models read structured text far more reliably than text copied out of a PDF. Copy-and-paste from a PDF often breaks lines in the middle of sentences, mixes two columns together, repeats the page header on every page and flattens tables into a run of numbers. Markdown keeps headings as headings and tables as rows and columns, which helps a model answer questions about the right section.
The same file works for every assistant: use it as PDF to Markdown for ChatGPT, or paste it into Claude. For retrieval systems (RAG) and search indexes, headings give natural places to split a document into chunks, and the page markers let answers cite the page they came from.
What Does the Output Look Like?
The file starts with front matter and marks each page with an HTML comment, which AI tools read but rendered Markdown hides:
---
title: "Annual Review 2026"
source: "annual-review.pdf"
pages: 4
---
<!-- page 1 -->
# Annual Review 2026
## IntroductionPictures are left out, page numbers and repeated headers and footers are removed, and words split by a hyphen at the end of a line are joined again. In our test, a 4-page report came out in about half a second as 4.6 KB of Markdown, with the character count and token estimate shown next to the download.
How Accurate Is the Token Estimate?
The estimate uses about 4 characters per token for English and fewer for other scripts, which is close to how current models count. Treat it as a guide for staying inside a model's context window, not an exact figure.
For plain words only, use PDF to Text, and for a spreadsheet of the tables, PDF to Excel.
Frequently asked questions
Is it free?
Yes, free with no sign-up, no watermark and no limit on files or pages.
Is my document sent to an AI service?
No. The conversion happens in your browser and nothing is uploaded, to us or anyone else. You choose where to paste or upload the Markdown afterwards.
Can it handle scanned PDFs?
Yes, scanned pages are read with text recognition (OCR) in your browser. Choose the languages under Scanned pages (OCR).
Does it work as PDF to Markdown for RAG?
Yes, the headings make natural chunk boundaries and the page comments let each chunk carry its page number. Split on ## headings, or on the page comments for page-sized chunks.
Should I use JSON instead?
Use PDF to JSON when a program needs each block as data; Markdown is better when a model reads the text. For a plain Markdown file without AI extras, use PDF to Markdown, or the PDF Converter for every other format.
Last updated: