Zhipu Deploys GLM 5.2 to All GLM Coding Plan Tiers With 1M-Token Context — API and MIT Open Weights Arriving Next Week, No Benchmarks Published

Zhipu AI rolled out GLM 5.2 to all four tiers of its GLM Coding Plan (Lite, Pro, Max, Team) on June 13, featuring a 1M-token context window and two thinking modes — High and Max, with Max recommended for complex coding — and a planned MIT open-weight release the following week. No independent benchmark results have been published at launch; the company positions it as superior to prior GLM versions on long-horizon coding but third-party verification is pending. The staged rollout continues China's accelerating open-source coding model competition, with GLM 5.2 targeting developers seeking permissively licensed alternatives to Claude Code and GPT-5.5 across the Asia-Pacific market.