In a previous post , I compared MinerU, Granite-Docling, and PaddleOCR-VL. Many commentors suggested I added their favorite parsers. So I did. And also added some new capabilities to differentiate the top models. Here is the full list of parser compared: MinerU 2.5 (1.2B VLM) Granite-Docling (258M VLM) PaddleOCR-VL (0.9B VLM) XBerg 1.0 (text-layer parser, CPU) HURIDOCS PDLA v0.0.35 (VGT layout model + Tesseract) LiteParse 2.11 (Tesseract based, CPU) Chandra (Datalab's OCR model)...