PDF Inspector

PDF Inspector is a fast Rust library that classifies PDFs and extracts structured text, bypassing OCR when unnecessary to optimize processing speed and costs.

PDF Inspector is a Rust library designed for efficient PDF classification and text extraction. It identifies whether documents are text-based, scanned, or mixed by sampling content streams, allowing for per-page routing decisions. The library parses documents once, enabling rapid extraction of text and structure into Markdown format, including tables, headings, and lists. It supports bindings for Python, Node.js, and browser WebAssembly. This tool is built to reduce latency and infrastructure costs by avoiding OCR for native-text documents while providing selective OCR capabilities for scanned pages.

PDF Inspector at a glance

Best for: Software engineer
Runs on: Web, Rust, Python, Node.js, CLI
Pricing: Free
Released: February 2026
Category: AI productivity tool
Website: github.com/firecrawl/pdf-inspector

Key features

Smart document classification for text-based, scanned, or mixed PDFs
Position-aware text extraction with font and coordinate data
Markdown conversion including tables, headings, and code blocks
Dual-mode table detection using drawing operations and text heuristics
Selective per-page OCR routing to minimize processing overhead
Browser WebAssembly support for local parsing without server round trips
Multi-column and reading order detection
CID font support for various character encodings

Use it for

Optimizing document processing pipelines for large-scale ingestion
Routing scanned documents to OCR while extracting native text directly
Converting financial documents and legal papers to structured Markdown
Parsing research papers for local analysis or RAG pipelines

Alternatives to PDF Inspector

Docling — Provides advanced PDF understanding and local execution for sensitive data.
Pagesift — Extracts web data with high memory efficiency and speed for AI agents.
docTR — Extracts text and layout details from documents using deep learning.
OpenDataLoader PDF — It ranks first in benchmarks for layout and table extraction without needing a GPU.
deepdoctection — Orchestrates layout analysis, OCR, and model fine-tuning in PyTorch pipelines.
RapidOCR — It converts PaddleOCR models to ONNX format for fast inference and deployment.
MiniCPM OCR — Efficient document parsing with correctness assessment and high throughput.
Xberg — Extracts documents, code, images, and audio using one native-Rust engine.
Sparrow — Extracts structured data locally from documents without cloud dependencies.
jscpd — A native Rust binary that runs fast without a Node.js runtime and integrates with AI agents.
MinerU — Parses complex documents and tables into structured markdown for agents.
CamLabs.ai — Verifies if redacted text can still be copied out of a PDF file.

All PDF Inspector alternatives