Summary
DeepSeek, a Chinese start-up, has launched its AI model, DeepSeek V3, which outperforms Meta and OpenAI's models in benchmark tests.
The DeepSeek V3 model was trained in about two months, with 671 billion parameters, at a cost of $5.58 million.
The development showcases the significant progress of Chinese AI companies despite challenges such as US sanctions on semiconductor access.
DeepSeek, a rising star in the Chinese AI industry, has made a major breakthrough with the launch of its powerful language model, DeepSeek V3. This new model has surpassed industry giants like Meta Platforms and OpenAI in key performance benchmarks, signaling a strong move by Chinese tech companies to compete in the global AI market.