Recently, SenseTime, in collaboration with the Shanghai AI Lab and other institutions, released the large language model InternLM-123B. In tests across 51 global evaluation sets, it ranked first in 12 categories, securing the second place globally in overall strength, surpassing some metrics of GPT-4. This achievement is largely attributed to SenseTime's robust capabilities in data processing and its extensive SenseCore computing power foundation. In terms of technological iteration speed, SenseTime is also ahead of its peers, upgrading InternLM-104B to 123B within just six months. It can be said that the strong performance of InternLM-123B has narrowed the gap between China and the rest of the world in the field of large models. If SenseTime can maintain this pace of technological advancement, it is likely to become the first domestic large model to reach the level of GPT-4.