Nougat
Neural OCR model by Meta for academic documents and mathematical expressions.
About
Nougat, Neural Optical Understanding for Academic Documents by Meta, converts academic PDFs into Markdown using a vision-encoder-decoder model. It is strong on mathematical expressions in LaTeX, tables, and scientific notation, which trip up conventional OCR, and it runs through a Python package or API. Inference uses a GPU. Released under the CC-BY-NC license.
Reviews (0)
Leave a Review
No reviews yet. Be the first to review!
Details
- Category
- OCR & Document Processing
- Price
- Free
- Platform
- Local/Desktop
- Difficulty
- Intermediate (3/5)
- License
- CC-BY-NC
- Minimum VRAM
- 6 GB
- Added
- Apr 3, 2026
Related Tools
Python library for extracting text, tables, and metadata from PDFs.
Python bindings for MuPDF library for fast PDF text and image extraction.
Python library for extracting tables from PDF files.
Ready-to-use OCR library supporting 80+ languages with simple Python API.
Turn-key OCR system for historical and non-Latin script documents.
Vision-language model based OCR toolkit by AI2 for document understanding.