NanoNets/docext
An on-premises document intelligence toolkit using vision-language models to extract structured information from PDFs and images.
Velocity · 7d
+4.6
★ / day
Trend
→steady
star history
Docext provides OCR-free document extraction and conversion powered by VLMs, enabling conversion of documents to structured markdown with semantic understanding. It includes a 3B parameter model (Nanonets-OCR-s) specifically trained for image-to-markdown conversion and offers a benchmarking leaderboard for document processing evaluation. The toolkit is used in RAG pipelines for unstructured data extraction.