Models
lift
datalab-to/lift
A 9B vision model for document intelligence that extracts schema-aligned JSON from PDFs and images. It ships with local vLLM and Hugging Face inference options, a CLI, and a schema-building studio.
Topics
Related projects
Dolphin9.1kDolphin is ByteDance's open document image parsing model, released with weights and inference code to convert PDFs and images into structured JSON and Markdown.
mineru83.4kAn AI-powered document parsing app that converts PDFs and Office files into clean LLM-ready Markdown and JSON.
OCRmyPDF35.3kA command-line tool that adds a searchable OCR text layer to scanned PDFs, with multilingual recognition, deskewing, PDF/A output, and multi-core batch processing.
PyMuPDF11.1kPyMuPDF is a high-performance Python library built on MuPDF for extracting, analyzing, converting, rendering, and editing PDF and other documents.
Parsr6.2kParsr is a document cleaning, parsing and extraction toolchain that turns PDFs, images, DOCX and EML files into structured JSON, Markdown, CSV or text, with a Dockerized API and optional GUI.
Hands-On-AI-Engineering4.1kA curated collection of hands-on AI projects covering agents, RAG, OCR and multimodal starter builds.