From e7d918da869001284f77bb209a89537b9468ff57 Mon Sep 17 00:00:00 2001
From: cong <274902531@qq.com>
Date: Thu, 16 Jul 2026 11:53:17 +0800
Subject: [PATCH] docs: clarify built-in PDF conversion limitations

---
 README.md | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/README.md b/README.md
index aa2f58bb8..db52dde0b 100644
--- a/README.md
+++ b/README.md
@@ -24,6 +24,9 @@ MarkItDown currently supports the conversion from:
 - EPubs
 - ... and more!
 
+> [!NOTE]
+> The built-in PDF converter extracts text and attempts to preserve tables when they can be detected. Because PDF files encode visual layout rather than semantic structure, it may not infer headings, lists, or other Markdown structure, and the result may be mostly plain text. For layout-aware extraction, use [Azure Content Understanding](#azure-content-understanding) or [Azure Document Intelligence](#azure-document-intelligence).
+
 ## Why Markdown?
 
 Markdown is extremely close to plain text, with minimal markup or formatting, but still