Zhipu Releases GLM-5.2 for Long-Horizon Tasks

Zhipu (Z.ai) released GLM-5.2, a new entry in its GLM model family positioned for long-horizon tasks — multi-step workflows requiring sustained reasoning and tool use over extended sessions. The release was announced via Hugging Face’s blog, one of the few RSS-trackable channels currently available for this lab. No CVE or security incident is associated with this item; it’s tracked here as a model-launch signal for teams evaluating Chinese-lab alternatives to OpenAI, Anthropic, and Google models.