Chinese AI startup Z.ai, also known as Zhupai AI, released its GLM-5.1 language model. The model features 754 billion parameters and was released under a permissive MIT license. This license allows for commercial use and modification of the software.
GLM-5.1 positions itself as an open-source competitor to models from OpenAI, Anthropic, and Google. The release signals intensifying competition in the AI sector from Chinese developers.
The model achieved a state-of-the-art score of 58.4 on the SWE-Bench Pro benchmark. This performance exceeds the 57.7 score of OpenAI’s GPT-5.4. It also surpasses the 57.3 score of Anthropic’s Claude Opus 4.6.