DeepSeek, the Chinese AI startup founded in 2023 by Liang Wengfen, has recently sparked interest in the tech world after unveiling its R1 AI model powered by V3. According to benchmarks, the AI model surpasses OpenAI's o1 reasoning model across math, science, and coding. More interestingly, the model was trained with $6 million, reportedly a fraction of the development cost for proprietary models.