Chinese AI lab DeepSeek has launched DeepSeek-R1, a reasoning model that claims to outperform OpenAI's o1 on benchmarks like AIME, MATH-500, and SWE-bench Verified. R1, featuring 671 billion parameters, is available on Hugging Face under an MIT license, allowing commercial use. However, it faces regulatory constraints in China, restricting its responses on sensitive topics. This development follows U.S. proposals for stricter export rules on AI technologies to China, raising concerns about the competition.
