Zhipu AI has released GLM-5.1, a globally leading open-source flagship model, with a core breakthrough in its capability for "long-term tasks." It can continuously and independently operate for over 8 hours without human intervention, propelling AI models into a new stage of autonomous operation on an "hourly" scale. The model excels in professional coding abilities, ranking third globally in comprehensive performance across multiple benchmark tests, leading among domestic and open-source models. It surpasses GPT-5.4 and Claude Opus 4.6 in the SWE-Bench Pro evaluation. In practical applications, GLM-5.1 has demonstrated robust long-term operational capabilities, such as building a Linux desktop system from scratch within 8 hours and enhancing vector database query throughput to 6.9 times that of the initial version through 655 rounds of autonomous iteration.
