China just dropped an open source OCR LLM with 1.7 billion parameters:
https://huggingface.co/rednote-hilab/dots.ocr
What does China gain by releasing so many models for free? What are they trying to accomplish?