PDF Inspector for Accurate PDF to Markdown Conversion

PDF Inspector analyzes each page before converting PDF to Markdown.

It preserves useful reading order, tables, headings, lists, links, and code, using selective OCR when needed.

Start Convert

1

Upload file to convert

Upload your file

Drag and drop or click to select

or from website / public file URL

0/3-Upgradeto 30

Privacy & Performance: File processing runs entirely in your browser, so files are not uploaded. Processing speed may vary depending on your device and browser environment.

How to Inspect and Convert a PDF to Markdown

PDF Inspector combines document analysis with conversion so each page follows the right extraction path.

The workflow runs locally for files selected from your device and produces Markdown you can preview before downloading.

1

Add Your PDF

Choose a PDF from your device or provide a public file URL, then add it to the conversion queue.

2

Inspect the Page Structure

Let PDF Inspector identify the document type, complex layouts, pages that need OCR, and possible font-encoding problems.

3

Convert with Selective OCR

Convert embedded text directly and, when Automatic OCR is enabled, route only the flagged scanned or image-based pages through OCR.

4

Review and Download Markdown

Preview the reconstructed headings, paragraphs, lists, tables, links, and code blocks, then download the result as a Markdown file.

Why PDFs Need Inspection Before Markdown Conversion

A PDF describes where content appears on a page, not how that content should read in Markdown.

PDF Inspector examines the signals hidden behind the visual page before deciding how to extract it.

One File Can Contain Different Page Types
A digital report may include selectable text, scanned appendices, image-only inserts, and mixed pages. Applying one extraction method to the entire file can omit content or waste time on unnecessary OCR.
Visual Order Is Not Reading Order
PDF text is often stored as positioned fragments. Without font and coordinate context, two-column papers, sidebars, captions, and right-to-left passages can be joined in the wrong sequence.
Tables May Not Exist as Data
Many PDFs draw a table from lines and independently placed words rather than storing rows and columns. Financial statements, footnotes, and tables continued on later pages need layout-aware reconstruction.
Image 1

Inside the PDF Inspector Conversion Pipeline

PDF Inspector loads the document once and shares that parse between analysis and extraction, keeping page decisions aligned with the resulting Markdown.

Instead of assuming every page is the same, it combines PDF classification, spatial text analysis, structure recovery, and optional OCR routing.

Use a lightweight sampling pass to group the file as TextBased, Scanned, ImageBased, or Mixed, then record a 0.0-1.0 confidence score and page-level OCR candidates. Inspection typically takes tens of milliseconds, although file and device complexity affect timing.

Use font metrics and X/Y placement to rebuild paragraphs in a more natural sequence, including newspaper-style columns, sequential column flow, and right-to-left text.

Infer H1-H4 headings from relative font sizes, recognize bullet, numbered, and lettered lists, detect monospace code, preserve bold and italic emphasis, link URLs, and separate pages cleanly.

Recover bordered tables from PDF drawing operations and borderless tables from repeated text alignment, with handling for financial layouts, attached footnotes, and continuations across pages.

Decode Type0 and Identity-H fonts through embedded ToUnicode CMaps, with support for UTF-16BE, UTF-8, and Latin-1 text while flagging pages whose mappings remain unreliable.

Why Use PDF Inspector for PDF to Markdown

Start with evidence about the PDF, then produce Markdown that is easier to verify, edit, and reuse.

Avoid Blanket OCR
Keep fast native text extraction for digital pages and reserve OCR for the pages that actually need it, reducing duplicate text and needless image recognition.
Keep Local Files Private
PDF Inspector runs its Rust parser as WebAssembly inside a Web Worker, so PDFs selected from your device are processed in the browser without a file upload or server round trip.
Preserve Complex Reading Flow
Position-aware extraction helps Markdown follow columns and right-to-left passages in a useful sequence instead of flattening every text fragment by its internal storage order.
Retain More Table Meaning
Drawing-based and alignment-based detection gives bordered and borderless tables a path to Markdown, including layouts that a plain text extractor would scatter across lines.
Create Reviewable Markdown
Headings, lists, code blocks, emphasis, links, tables, and page boundaries make the output easier to compare with the source before it enters documentation or retrieval systems.
Use a Lightweight First Pass
The core PDF Inspector parser is written in Rust, relies on lopdf for PDF parsing, and does not need an ML model for text-based pages; an OCR model is loaded only when OCR is enabled and a page is routed to it.

PDF Inspector and PDF to Markdown FAQ

Clear answers about document classification, structured extraction, OCR, privacy, and conversion limits.