Search
Meituan Releases LongCat-Flash-Thinking Inference Model, Achieving SOTA Performance in Multiple Aspects in Open Source
2025-09-22 /
Read about 0 minute
Author：小编
On September 22, 2025, Meituan's LongCat team officially released the efficient inference model LongCat-Flash-Thinking. While maintaining rapid inference capabilities, this model enhances reasoning abilities in fields such as logic, mathematics, coding, and intelligent agents, becoming the first large language model in China to combine both 'deep thinking + tool invocation' and 'informal + formal' reasoning capabilities. Currently, the model has been fully open-sourced on HuggingFace and Github.
-
C114 Communication Network -
Communication Home
7 X 24 Track global technological trends
