DeepSeek's first reasoning model R1-Lite-Preview turns heads, beating OpenAI o1 performance

The company's published results highlight its ability to handle a wide range of tasks, from complex mathematics to logic-based scenarios.