Zhipu AI open-sources GLM-5.2 coding model with 1M-token context and MIT licensing

Zhipu AI released GLM-5.2 on June 16 under MIT license, with model weights freely available. The 1-million-token context window — roughly 750,000 words of usable memory — and dual reasoning modes let developers optimize compute costs per request. Flat-rate pricing replaces per-token fees, removing a friction point for high-volume developer workflows.