Meet dots.ocr: A New 1.7B Vision-Language Model that Achieves SOTA Performance on Multilingual Document Parsing Leave a Comment / By / August 17, 2025