2026-04-11 20:43 UTC
#OpenDataLoader PDF โ #opensource #PDF parser for #AI & #accessibility ๐ #1 in benchmarks (0.90 overall), runs 100% locally, zero cloud, no GPU needed. #RAG #LLM #Python #NodeJS #Java
๐ Extracts #Markdown, #JSON (with bounding boxes) & #HTML from any PDF โ correct reading order, heading hierarchy, list & image detection with XY-Cut++ algorithm
๐งต๐#pdf
Replies (2)
-
@michabbb@social.vivaldi.net 2026-04-11 20:43
๐ Ranked #1 overall (0.90) across 200 real-world PDFs โ 0.93 table accuracy, 0.94 reading order in hybrid mode. Beats docling (0.86), marker (0.83), pymupdf4llm (0.57) โก Two modes: Fast local (0.05s/page, CPU-only) for standard PDFs + Hybrid mode (0.43s/page) for complex pages โ AI backend runs on your machine, zero cloud dependency
-
Are you aware that there is no conclusive reference here?