Data as of Jul 25, 2026 · Based on 2,718,867 AI responses across 9,511 prompts · See how Parse measures this
docTR is an open-source OCR library powered by PyTorch that provides state-of-the-art document text detection and recognition. It offers pretrained models and a simple API for extracting text from documents, supporting both research comparison and production deployment.
Parse Score
Sources
arxiv.org shapes more of what AI says about DocTR than any other source, at 100% of its citations.
