pdf-inspector.wiki is the bilingual community site and usage guide for pdf-inspector — the open-source Rust library that classifies PDFs and turns them into Markdown.

What this site is

pdf-inspector is Firecrawl's open-source Rust parser: it detects whether a PDF is text-based or scanned within milliseconds and converts native-text PDFs into clean, position-aware Markdown locally. This site does three things around it:

1. Online demo — drop a PDF right in your browser; everything runs locally and files never leave your device
2. Documentation — from getting started to all five bindings, organized as a proper guide
3. Benchmark breakdown — the official opendataloader-bench comparison explained honestly, limitations included

Our stance

• We don't copy or rewrite official code — every parse happens inside the official parser or a library you installed locally
• No exaggeration — benchmark numbers come from the official repo, and limitations (like no OCR in the browser build) are stated plainly
• Open source respected — pdf-inspector uses the MIT License; this site's content is free to share with attribution

About Firecrawl

Firecrawl is a data crawling and parsing platform built for AI workloads. pdf-inspector is the parsing engine of its hybrid OCR pipeline — and one of several libraries it maintains in the open.

Site features

🔍 Online parsing demo

The official WebAssembly build runs right on the home page: drop a PDF, get the type verdict and Markdown instantly — nothing ever uploaded.

📚 Bilingual guides

Getting started, API quick reference, and full pages for Node / Python / Rust / WebAssembly / CLI — in both English and Chinese.

📊 Benchmark breakdown

Complete data from the five-engine opendataloader-bench comparison, with an honest list of limitations to inform your selection.

🌐 Static & privacy-friendly

This is a fully static site: no accounts, no document collection. See the privacy policy.