mistral-ocr

Collection of PDF parsing libraries like AI based docling, claude, openai, gemini, meta's llama-vision, unstructured-io, and pdfminer, pymupdf, pdfplumber etc for efficient snapshot, text, table, and metadata extraction.

synthetic dataset generation workflow using local file resources for finetuning llms.

Turn Zotero PDFs into structured, source-linked Markdown with MinerU.

Extract data from images, pdf, invoices, receipts | Extract tables from pdf, images and convert to Excel/CSV | OCR complex pdfs, images.

Open, independent benchmark on document parsing APIs and frontier VLMs for AI agents and RAG: does a struck-out clause in a redlined contract PDF survive the parse? Open source code + open data. LlamaParse, Reducto, Datalab, Extend, Pulse, Mistral OCR, GPT-6 Astra, Claude Fable 5.1.

This is a collection of various document parsers and hands-on to construct structured data for your RAG applications.

ARCHIVED — first figmark (2026). Superseded by a new project at github.com/Ztein/figmark.

Multi-Agent System for Sell-Side Research Analysis. Powered by autonomous agent coordination, structure-preserving RAG, and first-class citation tracking.

Three modern PDF parsers stress-tested on a real European magazine. Editorial layouts, vertical text, multi-column TOCs. LlamaParse vs Mistral OCR vs Docling.

A minimal Mistral API wrapper to OCR images and PDFs with images embedded in the markdown result.

A python package with graphical user interface for processing images with the Mistral OCR API

Batch OCR of PDFs using PyMuPDF and Mistral OCR

A helper LLM app with RAG and Mistral OCR for very long reports like 10-Q, 10-K filed with the SEC.

Faithful image and PDF OCR to traceable Markdown with Mistral OCR.

Implemented an automation workflow that saves me >90% of time spent logging expenses!

A Python helper for extracting text from PDFs and images using Mistral OCR

Document Intelligence with Mistral OCR 4.0 in Azure AI Foundry via direct API calls and Entra ID auths.

Prototype Mistral OCR pipeline. Upload your French PDF and download in English. Have it look the same. A dead simple solution for the travelling elderly and small businesses working in another language — anyone not tech savvy. Future port to Next.js.

Multi-engine OCR suite with Tesseract, EasyOCR, PaddleOCR, Docling, and Mistral OCR — desktop apps, web UIs, Jupyter notebooks, and a unified FastAPI hub for document text extraction.

Convert image-only picture-book PDFs into fixed-layout EPUB3 + editable PPTX — Mistral OCR 4 for text+layout, LaMa inpainting to erase baked-in text, deterministic fit-to-box typesetting.