HONG KONG -- Chinese AI startup DeepSeek quietly released an update to its V3 large language model on Tuesday evening, significantly enhancing its reasoning capabilities and further escalating competition with U.S. rival OpenAI.
Dubbed DeepSeek-V3-0324, the updated model has improved significantly over its predecessor, especially in advanced math and coding, according to the company's post on Hugging Face, a U.S. open-source AI community.
