MonkeyOCRv2: A New Document AI Foundation Model

Researchers introduced MonkeyOCRv2, a vision-text foundation model for Document AI featuring a document-native vision encoder and trained on the massive 113-million-image MonkeyDoc v2 dataset.

2026-07-14 ~ 2026-07-15 · 2 related posts

• MonkeyOCRv2: A Foundation Model for Document AI — _reachsumit · 2026-07-14
• MonkeyOCRv2: Document AI Vision Model — VLRLab-OCR · 2026-07-15