allenai/olmocr
Summary
olmOCR is a toolkit from the Allen Institute's AllenNLP team for converting PDFs and other image-based documents into clean, readable plain text using vision-language models. It supports local GPU inference, remote inference via vLLM, Beaker cluster execution, and Docker images, plus a comprehensive installation and benchmarking guide. This makes it relevant for automating large-scale document digitization workflows in research and SMB contexts.