Firecrawl Releases pdf-inspector: Fast PDF Classification and Markdown Extraction in Rust

X postNicolas CamaraTool

Firecrawl has announced pdf-inspector, an open-source Rust library designed to classify PDF types rapidly and convert them to clean Markdown without requiring OCR. The tool targets AI agents and automated pipelines, processing text-based documents locally in milliseconds.

Key points

  1. Classifies documents into TextBased, Scanned, ImageBased, or Mixed categories in roughly 10 to 50 ms.
  2. Extracts text with positional awareness and converts it directly to Markdown, supporting tables and multi-column layouts.
  3. Supports selective OCR for scanned pages using PP-OCRv6 Small, bypassing heavy runtimes for clean text pages.
  4. Provides native Rust APIs alongside bindings for Python, Node.js, and WebAssembly for in-browser execution.
  5. Outperformed comparison parsers in speed, reading order, and table structure tests on the 200-document opendataloader-bench corpus.

Why it matters

AI workflows and parsing pipelines frequently encounter bottlenecks and infrastructure costs when routing every document through full OCR. Pre-filtering and handling digital-native PDFs locally removes significant latency for the large share of PDFs that already contain structured text.

Analysis

The project serves as an efficient front-end filter for ingestion pipelines. By inspecting content streams and drawing commands directly in Rust, it determines whether a document requires heavy optical processing or can be converted to Markdown on the spot, streamlining agent interactions with financial, academic, and business documents.

Original post

Nicolas Camara

@nickscamara_ · 31 July 2026

we built pdf-inspector so agents can process PDFs without waiting on OCR. it classifies any PDF in ~20ms and extracts clean markdown locally → 200 PDFs processed in 2.8s → top quality in extracting tables + graphs → built in rust → open source t.co/Wjp9kpTHXJ t.co/giWEddU6ug

10.9k likes294 replies20.6k bookmarks2.7m views

View on X
Selected replies (5)
  • @rfltinoco @nickscamara_ @nrtawari Hey @nickscamara_, we just saw that come true. A document that took 45 min to process with Docling took 1.5 seconds with pdf-inspector. Crazy stuff!
  • @homsiT @sirfloriank @nickscamara_ seems to not support images, which are pretty important... (most pdfs added to Reader have images)
  • @bhadravaghela7 @nickscamara_ Try: t.co/bLYdZ9ToGW
  • @aithura_com @nickscamara_ Impressive! 👌🏽
  • @dcoolisek @nickscamara_ Apparently we all decided PDFs need fixing 😅 You built it for agents, we built Bubu PDF for humans on Mac. t.co/sbafj5SZx5

Mentioned links

Links and tools from this entry

Summary generated by AI (Gemini 3.8 Flash, 27 September 2026). It may contain errors – the original source is authoritative.

Related entries

Similar in content, found by the AI search.

  • Link:docling.ai

    Docling: Open-Source Document Intelligence for AI and RAG

    Docling is an open-source tool hosted by LF AI & Data that converts complex documents like PDFs, Office files, and images into structured data for AI applications. It operates locally and fully offline, preserves layouts and table structures, and provides interfaces for Python, CLI, REST APIs, and MCP.

    AI & LLMs· Tool

  • Repository:unclecode/crawl4ai

    Crawl4AI: Open-Source LLM-Friendly Web Crawler and Scraper

    Crawl4AI is an open-source Python web crawler and scraper designed to transform web pages into clean, LLM-ready Markdown for RAG setups, AI agents, and data pipelines.

    81.2kstarsPython

    AI & LLMs· Tool

  • Repository:docling-project/docling

    Docling: Document Parsing for Generative AI

    Docling is an open-source Python tool designed to parse documents across multiple formats—such as PDF, office files, audio, and HTML—to prepare them for generative AI workflows.

    66kstarsPython

    AI & LLMs· Tool

Back to the knowledge base

Let's talk about your project

Get in touch

Locations

  • Mattersburg
    Johann Nepomuk Bergerstraße 7/2/14
    7210 Mattersburg, Austria
  • Vienna
    Ungargasse 64-66/3/404
    1030 Wien, Austria

Parts of this content were created with the assistance of AI.