A few days ago, while researching text-to-video models, I discovered something that differed significantly from my original understanding: in the directions of text-to-image and text-to-video, Chinese companies' presence in the open source community is far weaker than in language models.

I had always assumed Chinese models dominated across all...