Comment by dryarzeg

Then you should consider alternative LLM API providers, who are not based in China but host the same (or roughly the same, depending on the quantization and other deployment specifics) models as your "Chinese AI services".