ChatGPT ChatGPT & SiteToMarkdown

Give ChatGPT the Full Context of Your Documentation

Don't rely on ChatGPT's training cutoff or slow web browsing. Convert entire documentation sites into a single, optimized Markdown file that you can upload to ChatGPT Projects and Custom GPTs.

Convert a Site to Markdown See how it works
Adding Markdown files to a ChatGPT Project

Why ChatGPT needs SiteToMarkdown

ChatGPT's file upload feature is powerful, allowing for synthesis, transformation, and extraction of data. However, it has strict limits:

  • File Count Limits: You are limited to 20 files per Project (Plus) or 40 files (Team). You can't upload a documentation site page-by-page.
  • Browsing is Flaky: Asking ChatGPT to "read the docs" via Bing often fails on complex, JS-heavy sites or only reads the homepage.
  • Training Cutoffs: ChatGPT doesn't know about the API changes released last week unless you provide the files.

The SiteToMarkdown Advantage

We crawl every page of a documentation site and merge it into a single Markdown file.

  • Fits the Limits: One file takes up only 1 of your 20 available slots.
  • Token Efficient: Markdown is stripped of HTML bloat, maximizing the 2M token limit per file.
  • Persistent Context: Upload once to a Project, and the knowledge is available for every new chat.

Build a Knowledge Base for ChatGPT

Follow this workflow to give ChatGPT expert-level knowledge of any library or framework.

1

Generate the Markdown

Paste the URL of the documentation (e.g., stripe.com/docs) into SiteToMarkdown. We'll fetch all subpages and combine them into one clean file.

2

Upload to ChatGPT

Open ChatGPT and go to Projects or create a Custom GPT. Upload the Markdown file to the knowledge base.

3

Chat with Context

Start a new chat within that Project. ChatGPT will now reference the full documentation to answer questions, write code, and debug issues.