DeepSeek-R1 official update statement: Think deeper, reason stronger.

DeepSeek-R1-0528 still uses the DeepSeek V3 Base model released in December 2024 as a base, but more computing power has been invested in the post-training process, significantly enhancing the model's depth of thought and reasoning abilities. The updated R1 model has achieved outstanding results in various benchmark evaluations such as mathematics, programming, and general logic, outperforming all other models currently available domestically. Overall, it is now approaching the performance level of other top international models, such as o3 and Gemini-2.5-Pro.