GLM 5.3
Z.AI · text → text
GLM-5.3 is a coding and agentic reasoning model developed by Z.AI, designed for complex software engineering, long-horizon agent tasks, and vulnerability analysis. Built on GLM-5.2, it incorporates further post-training enhancements to improve coding capabilities, task execution, and token efficiency. The version currently available is an internal preview of GLM-5.3, with a TPM of approximately 1–2 million. As the model is still in testing, service stability is not guaranteed and it is not recommended for production use. We plan to add additional resources and capacity over the coming week to further improve availability and stability.
Input$1.13 /M
Output$3.94 /M