【Tech 24H】Baichuan AI released its enhanced large language model Baichuan-M2, which scored 60.1 on the HealthBench medical evaluation benchmark. Despite its compact 32-Byte size, it outperformed OpenAI’s latest open-source model gpt-oss120b (scored 57.6) as well as other leading global models including Qwen3-235B, Deepseek R1, and Kimi K2.