Yunzhisheng announced the launch of the 2.0 version of the Shanhai large model, with a parameter scale reaching trillions. The actual performance surpassed GPT-4 in the C-Eval global large model comprehensive evaluation, ranking among the top three with a score of 70. The model team enriched the model's knowledge base by utilizing textbooks, literature, and encyclopedic materials, achieving breakthroughs particularly in the medical field.