# pdfmux blog
> Engineering notes on PDF extraction, OCR, and LLM document pipelines.
## Posts
- [Docling vs LlamaParse: Which PDF Parser to Pick for RAG (2026)](https://pdfmux.com/blog/docling-vs-llamaparse/): Docling vs LlamaParse for RAG PDF extraction: open-source vs hosted, setup, tables, OCR, output, self-hosting, pricing, and when to pick each.
- [Mistral OCR for PDF-to-Markdown Conversion (2026 Review)](https://pdfmux.com/blog/mistral-ocr-pdf-markdown/): A developer look at Mistral OCR for PDF-to-Markdown: what it extracts, pricing per page, language support, limits, and when a local tool fits better.
- [pdfmux vs Google Document AI: PDF Extraction for LLM and RAG Pipelines (2026)](https://pdfmux.com/blog/pdfmux-vs-google-document-ai/): How pdfmux and Google Document AI compare for PDF extraction: output format, pricing, setup, data residency, and which fits LLM and RAG pipelines in 2026.
- [PyMuPDF vs pdfplumber (2026): the speed-vs-license tradeoff, benchmarked](https://pdfmux.com/blog/pymupdf-vs-pdfplumber/): PyMuPDF is 10x faster but AGPL-licensed. pdfplumber is MIT but slow. Benchmarks on 1,422 real pages, TEDS table scores, and a third option that beats both.
- [Unstructured vs LlamaParse: Which PDF Parser to Pick for RAG (2026)](https://pdfmux.com/blog/unstructured-vs-llamaparse/): Unstructured vs LlamaParse for RAG PDF extraction: open-source vs hosted, setup, tables, OCR, output, self-hosting, pricing, and when to pick each.
- [Arabic PDF OCR in Python: extract right-to-left text from scanned documents](https://pdfmux.com/blog/arabic-pdf-ocr/): How to OCR Arabic PDFs in Python. Handle right-to-left order, bidirectional text, and ligature reshaping with Tesseract, and when to reach for multimodal OCR.
- [Multilingual PDF to Markdown: extract non-English documents in Python](https://pdfmux.com/blog/multilingual-pdf-to-markdown/): Extract text from PDFs in 100+ languages and convert to clean Markdown. Detect the language, pick the right OCR engine, and handle mixed-script pages in Python.
- [Docling vs Marker: Which Open-Source PDF-to-Markdown Library to Pick (2026)](https://pdfmux.com/blog/docling-vs-marker/): A developer comparison of Docling (IBM) and Marker (VikParuchuri) for PDF-to-Markdown extraction: install, speed, table and equation accuracy, GPU needs, OCR, license, and
- [We ran pdfmux on Tesla 10-Ks, Uber S-1s, and Supreme Court opinions. Here is what happened.](https://pdfmux.com/blog/real-world-pdf-benchmark/): Real-world benchmark of pdfmux vs PyMuPDF and pymupdf4llm across 11 public documents — 1,422 pages of SEC filings, academic papers, legal opinions, and government reports.
- [PDF extraction routing in Python: pick the right parser per page](https://pdfmux.com/blog/pdf-extraction-routing-python/): How to route each PDF page to the best extractor in Python using text density, font embedding, and scan detection — so you stop OCR-ing pages that already have clean text.
- [How to process large PDFs in Python without running out of memory](https://pdfmux.com/blog/process-large-pdf-python/): Extract text from 1,000+ page PDFs in Python without OOM crashes. Compares load-everything, page iteration, and streaming approaches with memory benchmarks and code.
- [OCR PDF extraction in Python: extract text from scanned PDFs (2026 guide)](https://pdfmux.com/blog/ocr-pdf-extraction-python/): How to extract text from scanned and image-based PDFs in Python using OCR. Compares pytesseract, EasyOCR, Surya, and pdfmux with code examples and accuracy benchmarks.
- [Extract text from multi-column PDFs in Python (in the right reading order)](https://pdfmux.com/blog/multi-column-pdf-extraction-python/): Multi-column PDFs scramble text in naive extractors. How reading order breaks, how to detect column layout in Python, and how to get clean output.
- [Extract PDF metadata in Python: document info, XMP, and what pipelines need](https://pdfmux.com/blog/pdf-metadata-extraction-python/): Read PDF metadata in Python: the /Info dictionary, XMP, page count, and the extraction metadata (confidence, extractor) that actually matters for AI pipelines.
- [pdfmux vs MarkItDown — specialist or universal?](https://pdfmux.com/blog/compare/pdfmux-vs-markitdown/): MarkItDown is Microsoft's 153K-star universal-to-markdown converter. pdfmux is a PDF specialist with per-page confidence + audit manifest. Honest tradeoff, side-by-side.
- [PDF to Excel in Python: Convert Tables to Multi-Sheet XLSX](https://pdfmux.com/blog/pdf-to-excel-python/): Convert PDF tables to a multi-sheet Excel .xlsx in Python: extract structured JSON with pdfmux, write each table to its own sheet with pandas + openpyxl, fix number types.
- [pdfmux vs PyMuPDF vs marker vs docling vs pdfplumber: 200-PDF benchmark](https://pdfmux.com/blog/pdfmux-vs-pymupdf-vs-marker-vs-docling/): Head-to-head benchmark of 5 Python PDF extraction libraries on 200 real-world PDFs. Scores, runnable code for each tool, license traps, and which to pick for RAG.
- [Extract data from bank statements (PDF) in Python: a 2026 guide](https://pdfmux.com/blog/extract-bank-statement-data-python/): Parse bank statement PDFs into structured transactions in Python: multi-page tables, running-balance reconciliation, scanned OCR, and the code to do it.
- [Extract text from password-protected (encrypted) PDFs in Python](https://pdfmux.com/blog/extract-text-encrypted-pdf-python/): Decrypt and extract text from password-protected PDFs in Python with pikepdf and PyMuPDF: user vs owner passwords, AES-256, batch decryption, and clean text.
- [PDF to CSV in Python: 4 Methods (Tables to Clean CSV)](https://pdfmux.com/blog/pdf-to-csv-python/): Convert PDF tables to CSV in Python with pdfplumber, camelot, tabula-py, and pdfmux. Runnable code examples, a comparison table, and common pitfalls to avoid.
- [PDF to Markdown for RAG pipelines: the complete guide](https://pdfmux.com/blog/pdf-to-markdown-for-rag/): Convert PDFs to clean Markdown for RAG and LLM pipelines: chunking, tables, OCR, confidence scoring, and production patterns. Updated June 2026.
- [pdfmux confidence scoring explained: how we measure extraction quality](https://pdfmux.com/blog/pdfmux-confidence-scoring-explained/): A deep dive into pdfmux's five-check confidence scoring system — how it measures PDF extraction quality per page and routes bad pages to better extractors.
- [pdfmux vs Chandra — model or orchestrator?](https://pdfmux.com/blog/compare/pdfmux-vs-chandra/): Chandra is Datalab's newest VLM for PDF extraction — SOTA accuracy, SOC 2 Type 2, OpenRAIL-M model license. pdfmux is an orchestrator with an audit manifest. Honest
- [How to extract embedded images and figures from a PDF in Python](https://pdfmux.com/blog/extract-images-from-pdf-python/): Extract embedded images and figures from PDFs in Python with PyMuPDF, pdfplumber, pdf2image, and pdfmux. Code for raw bytes, bounding boxes, CMYK, and dedup.
- [pdfmux vs Azure Document Intelligence: cost, privacy, and accuracy in 2026](https://pdfmux.com/blog/pdfmux-vs-azure-document-intelligence/): Honest 2026 comparison of pdfmux and Azure AI Document Intelligence on accuracy, cost, privacy, and prebuilt models. Where each one wins, with the cost math.
- [How to extract tables from PDF in Python (3 methods compared)](https://pdfmux.com/blog/extract-tables-from-pdf-python/): Extract tables from PDF files in Python using pdfmux, PyMuPDF, and Docling. Code examples, accuracy benchmarks, and the best approach for each use case.
- [pdfmux vs LiteParse — library or product?](https://pdfmux.com/blog/compare/pdfmux-vs-liteparse/): LiteParse is a 8.4K-star OSS PDF parser from LlamaIndex. pdfmux is a product with audit-correctness + Cloud quotas. Honest tradeoff, side-by-side.
- [Best PDF extraction library for Python in 2026 (benchmarked)](https://pdfmux.com/blog/best-pdf-extraction-library-python/): We benchmarked 8 Python PDF libraries on 200 real-world PDFs. Ranked by table accuracy, reading order, and heading structure. Updated May 2026.
- [PDF extraction with Node.js in 2026: libraries, benchmarks, and the pdfmux MCP path](https://pdfmux.com/blog/pdf-extraction-with-nodejs/): Extract text, tables, and structured data from PDFs in Node.js. Library comparison (pdf-parse, unpdf, pdf2json), benchmarks, code, and the pdfmux MCP path.
- [PDF to JSON for LLM pipelines: structured extraction patterns in 2026](https://pdfmux.com/blog/pdf-to-json-for-llm-pipelines/): PDF to JSON for LLM tool use, function calling, and structured outputs. 3-layer schema, bounding boxes, confidence scoring, and the pdfmux pattern.
- [PDF extraction without GPU or API keys: how pdfmux does it](https://pdfmux.com/blog/pdf-extraction-without-gpu/): How pdfmux achieves 99% of AI-powered extraction accuracy with zero GPU, zero API keys, and zero cost per page. Architecture explained.
- [How to detect if a PDF is scanned (and what to do about it)](https://pdfmux.com/blog/detect-scanned-pdf-python/): Three reliable ways to detect a scanned PDF in Python, plus a routing pattern that runs OCR only on the pages that need it. With code and benchmarks.
- [PDF chunking strategies for RAG: page, heading, and semantic compared](https://pdfmux.com/blog/pdf-chunking-strategies-rag/): Four chunking strategies for PDF-backed RAG pipelines, benchmarked on retrieval recall. With code, a comparison table, and a recommendation by document type.
- [Use pdfmux inside Claude Code (and Claude Desktop, Cursor, any MCP agent)](https://pdfmux.com/blog/pdfmux-for-claude-code/): Two-minute setup to give Claude Code, Claude Desktop, and Cursor any-PDF parsing via the pdfmux MCP server. Local, offline, no API keys.
- [Loading PDFs into LlamaIndex: SimpleDirectoryReader, custom parsers, and confidence-aware ingestion](https://pdfmux.com/blog/loading-pdfs-into-llamaindex/): Three ways to load PDFs into LlamaIndex in 2026: SimpleDirectoryReader, PDFReader, and a custom file_extractor. Plus the pdfmux integration that gives you a confidence score
- [pdfmux vs AWS Textract: when self-hosted PDF extraction beats the cloud](https://pdfmux.com/blog/pdfmux-vs-aws-textract/): Honest 2026 comparison of pdfmux and AWS Textract on accuracy, cost, privacy, and integration. Where Textract still wins, and where the cost math has flipped against it.
- [I benchmarked every PDF-to-Markdown tool. Then I built a router.](https://pdfmux.com/blog/benchmarking-pdf-extractors/): No single PDF extraction tool wins at every category. I tested PyMuPDF, Docling, Marker, Surya, Mistral OCR, Gemma 4, and Gemini Flash across digital, table-heavy, scanned,
- [We ran pdfmux on 433 of our own customer's PDFs. The first run failed silently.](https://pdfmux.com/blog/when-our-own-pdf-extractor-failed-silently/): Run pdfmux on the same PDFs your current extractor processed, then diff the manifests. If pdfmux's confidence is low on documents your tool reported as successful, those are
- [PDF extraction with LangChain: loaders, splitters, and the pdfmux integration](https://pdfmux.com/blog/pdf-extraction-with-langchain/): How to load PDFs into LangChain. PyPDFLoader, UnstructuredPDFLoader, PyMuPDFLoader, and the pdfmux loader compared on speed, table fidelity, and cost.
- [pdfmux 1.6 — three new backends, smart cache, streaming, watch mode](https://pdfmux.com/blog/pdfmux-1-6-released/): What shipped in pdfmux 1.6.0: Mistral OCR, Marker, Gemma 4, hash-keyed result cache, NDJSON streaming, configuration profiles, watch mode, cost prediction, diff, retry with
- [Batch PDF processing in Python: handle 10,000 PDFs without a crash](https://pdfmux.com/blog/batch-pdf-processing-python/): Process thousands of PDFs in Python: memory-safe batching, parallel workers, and error recovery. Benchmarked patterns for production PDF pipelines.
- [How to extract data from fillable PDF forms in Python](https://pdfmux.com/blog/pdf-form-data-extraction-python/): Extract form data from PDFs in Python: AcroForms, XFA, and mixed docs covered. Code for pypdf, pdfrw, pdfmux with real-world examples.
- [pdfmux vs LlamaParse vs Docling vs Unstructured: Which PDF extractor for RAG in 2026?](https://pdfmux.com/blog/pdfmux-vs-llamaparse-vs-docling-vs-unstructured-2026/): Direct, benchmark-backed comparison of pdfmux, LlamaParse, Docling, and Unstructured for RAG pipelines in 2026 — accuracy scores, pricing, install size, GPU requirements,
- [Extract invoice data from PDFs in Python: a complete guide for AP automation](https://pdfmux.com/blog/pdf-invoice-extraction-python/): How to extract invoice data from PDFs in Python. Benchmarks on 1,200 real invoices, production code for field extraction, and a working AP automation pipeline.
- [pdfmux vs LlamaParse: accuracy, cost, and privacy compared (2026)](https://pdfmux.com/blog/pdfmux-vs-llamaparse/): Head-to-head: pdfmux versus LlamaParse on accuracy, cost, and privacy. Which PDF extractor fits your RAG pipeline in 2026 — and when does the cost math flip?
- [How to Extract Data from Arabic PDFs: A Complete Guide for GCC Logistics](https://pdfmux.com/blog/arabic-pdf-extraction-gcc-logistics/): Extract bilingual Arabic and English data from Bills of Lading, commercial invoices, and customs forms. Production code, benchmark data, and compliance notes for UAE and
- [Running PDF Extraction Locally with Gemma 4: Zero API Costs, Full Privacy](https://pdfmux.com/blog/local-pdf-extraction-gemma-4/): Replace cloud LLM extraction with Gemma 4 running locally via Ollama. Zero API costs, full privacy, 140+ language OCR. Hardware requirements, benchmark numbers, and setup
- [PDFMux MCP Server: Intelligent PDF Extraction Inside Claude, Cursor, and Windsurf](https://pdfmux.com/blog/pdfmux-mcp-server-claude-cursor/): The only PDF MCP server with confidence scoring, multi-method extraction, and structured output. Install in 30 seconds with npx -y pdfmux. Works in Claude Desktop, Cursor,
- [The LiteLLM supply chain attack: what every AI engineer should learn from it](https://pdfmux.com/blog/litellm-security-lessons/): LiteLLM, the most popular open-source LLM proxy, was compromised via a poisoned security scanner in its CI/CD pipeline. Here is what happened, why it matters, and what it
- [PDF data extraction for AI agents: the missing infrastructure layer](https://pdfmux.com/blog/pdf-data-extraction-for-ai-agents/): AI agents need to read PDFs to do real work — invoices, contracts, reports. But most extraction tools return text with no quality signal. Here is the pattern emerging across
- [pdfmux vs OpenDataLoader: why an orchestrator beats a single extractor for RAG](https://pdfmux.com/blog/pdfmux-vs-opendataloader/): OpenDataLoader PDF v2.0 tops benchmarks at 0.90 accuracy — but production RAG pipelines need more than one extractor. How pdfmux uses OpenDataLoader alongside PyMuPDF,
- [PDF extraction for RAG pipelines: from messy PDFs to accurate retrieval](https://pdfmux.com/blog/pdf-extraction-for-rag-pipeline/): Build a production RAG pipeline with clean PDF extraction. Covers pdfmux + LangChain + ChromaDB with code, chunking strategies, and how extraction quality directly impacts
- [pdfmux vs Kreuzberg: PDF extraction comparison (2026 benchmark)](https://pdfmux.com/blog/pdfmux-vs-kreuzberg/): Head-to-head comparison of pdfmux and Kreuzberg for Python PDF extraction. Benchmark results on 200 PDFs, code examples, feature comparison, and which to use for RAG
- [Best AWS Textract Alternatives in 2026](https://pdfmux.com/blog/alternatives/aws-textract-alternatives/): Looking for AWS Textract alternatives? Compare the top PDF extraction tools that work without cloud dependencies or per-page pricing.
- [Best Google Document AI Alternatives in 2026](https://pdfmux.com/blog/alternatives/google-document-ai-alternatives/): Looking for Google Document AI alternatives? Compare local and cloud PDF extraction tools with lower cost and complexity.
- [Best LlamaParse Alternatives in 2026](https://pdfmux.com/blog/alternatives/llamaparse-alternatives/): Looking for LlamaParse alternatives? Compare the top PDF extraction tools that run locally without cloud dependencies.
- [Best PyMuPDF Alternatives in 2026](https://pdfmux.com/blog/alternatives/pymupdf-alternatives/): Looking for PyMuPDF alternatives? Compare the top PDF extraction tools including pdfmux, pdfplumber, Marker, Docling, and more.
- [Best AWS Textract Alternatives in 2026](https://pdfmux.com/blog/alternatives/textract-alternatives/): Looking for AWS Textract alternatives? Compare the top PDF extraction tools that work without cloud dependencies or per-page pricing.
- [Best Unstructured.io Alternatives in 2026](https://pdfmux.com/blog/alternatives/unstructured-alternatives/): Looking for Unstructured.io alternatives? Compare lighter, faster PDF extraction tools for your document processing pipeline.
- [pdfmux vs AWS Textract: PDF Extraction Compared](https://pdfmux.com/blog/compare/pdfmux-vs-aws-textract/): Side-by-side comparison of pdfmux and AWS Textract — accuracy, cost, latency, operational complexity, and the tradeoffs that matter in production.
- [pdfmux vs Docling: PDF Extraction Compared](https://pdfmux.com/blog/compare/pdfmux-vs-docling/): Compare pdfmux and Docling for PDF text extraction. Features, benchmarks, pricing, and when to use each.
- [pdfmux vs Google Document AI: PDF Extraction Compared](https://pdfmux.com/blog/compare/pdfmux-vs-google-document-ai/): Compare pdfmux and Google Document AI for PDF text extraction. Features, benchmarks, pricing, and when to use each.
- [pdfmux vs LlamaParse: PDF Extraction Compared](https://pdfmux.com/blog/compare/pdfmux-vs-llamaparse/): Compare pdfmux and LlamaParse for PDF text extraction. Features, benchmarks, pricing, and when to use each.
- [pdfmux vs Marker: PDF Extraction Compared](https://pdfmux.com/blog/compare/pdfmux-vs-marker/): Compare pdfmux and Marker for PDF text extraction. Features, benchmarks, pricing, and when to use each.
- [pdfmux vs pdfplumber: PDF Extraction Compared](https://pdfmux.com/blog/compare/pdfmux-vs-pdfplumber/): Compare pdfmux and pdfplumber for PDF text extraction. Features, benchmarks, pricing, and when to use each.
- [pdfmux vs PyMuPDF: PDF Extraction Compared](https://pdfmux.com/blog/compare/pdfmux-vs-pymupdf/): Compare pdfmux and PyMuPDF for PDF text extraction. Features, benchmarks, pricing, and when to use each.
- [pdfmux vs pymupdf4llm: PDF Extraction Compared](https://pdfmux.com/blog/compare/pdfmux-vs-pymupdf4llm/): Compare pdfmux and pymupdf4llm for PDF text extraction. Features, benchmarks, pricing, and when to use each.
- [pdfmux vs Apache Tika: PDF Extraction Compared](https://pdfmux.com/blog/compare/pdfmux-vs-tika/): Compare pdfmux and Apache Tika for PDF text extraction. Features, benchmarks, pricing, and when to use each.
- [pdfmux vs Unstructured: PDF Extraction Compared](https://pdfmux.com/blog/compare/pdfmux-vs-unstructured/): Compare pdfmux and Unstructured for PDF text extraction. Features, benchmarks, pricing, and when to use each.
- [What Is Document Ingestion? Definition and Guide](https://pdfmux.com/blog/glossary/what-is-document-ingestion/): Document ingestion is the process of loading, extracting, and preparing documents for downstream processing. A complete guide for developers.
- [What Is Document Intelligence? Definition and Guide](https://pdfmux.com/blog/glossary/what-is-document-intelligence/): Document intelligence uses AI to understand, classify, and extract structured data from documents. A complete guide for developers.
- [What Is Layout Analysis? Definition and Guide](https://pdfmux.com/blog/glossary/what-is-layout-analysis/): Layout analysis detects the spatial structure of document pages — columns, headers, tables, and reading order. A complete guide for developers.
- [What Is OCR? Definition and Guide](https://pdfmux.com/blog/glossary/what-is-ocr/): OCR (Optical Character Recognition) converts images of text into machine-readable characters. A complete guide for developers.
- [What Is PDF Extraction? Definition and Guide](https://pdfmux.com/blog/glossary/what-is-pdf-extraction/): PDF extraction is the process of pulling text, tables, images, and metadata from PDF files programmatically. A complete guide for developers.
- [What Is a RAG Pipeline? Definition and Guide](https://pdfmux.com/blog/glossary/what-is-rag-pipeline/): A RAG pipeline (Retrieval-Augmented Generation) combines document retrieval with AI generation for accurate, grounded answers. A complete guide for developers.
- [What Is PDF Parsing? Definition and Guide](https://pdfmux.com/blog/glossary/what-is-pdf-parsing/): PDF parsing is the process of reading and interpreting the internal structure of PDF files. A complete guide for developers.
- [What Is Text Chunking? Definition and Guide](https://pdfmux.com/blog/glossary/what-is-text-chunking/): Text chunking splits documents into appropriately sized segments for embedding, retrieval, and AI processing. A complete guide for developers.
- [What Is a Vector Embedding? Definition and Guide](https://pdfmux.com/blog/glossary/what-is-vector-embedding/): Vector embeddings convert text into numerical representations that capture semantic meaning. A complete guide for developers.
- [What Is Table Extraction? Definition and Guide](https://pdfmux.com/blog/glossary/what-is-table-extraction/): Table extraction is the process of detecting and converting tables in documents into structured data. A complete guide for developers.
- [Which PDF extractor should you actually use in 2026?](https://pdfmux.com/blog/pdf-extractor-comparison-2026/): There are now 7+ serious PDF extraction tools — OpenDataLoader, Docling, Marker, MinerU, pymupdf4llm, MarkItDown, pdfmux, and more. Here is when to use each one, with real
- [What "self-healing" PDF extraction actually looks like](https://pdfmux.com/blog/self-healing-pdf-extraction/): Most PDF extractors run once and hope for the best. pdfmux extracts, audits every page with 5 quality checks, and re-extracts failures automatically — here is exactly how
- [Which PDF extractor should you use? An honest guide.](https://pdfmux.com/blog/which-pdf-extractor-should-you-use/): I maintain a tool that uses PyMuPDF, Docling, Surya, RapidOCR, and Gemini Flash internally. Here is when each one wins, when each one fails, and how to pick the right one
- [How to give your AI agent the ability to read any PDF](https://pdfmux.com/blog/mcp-server-pdf-ai-agent/): There are 18,000+ MCP servers but most PDF ones just wrap basic text extraction. Here is how to add production-grade PDF processing to Claude, Cursor, or any MCP-compatible
## About
- [pdfmux](https://pdfmux.com)
- [GitHub](https://github.com/NameetP/pdfmux)