A Model-Level 429 From Zhipu Took GLM Flash Models Offline. TheRouter Now Scopes Cooldowns Per Model and Gives Them a Second Route.
On 2026-10-02 Zhipu's API returned a model-level rate-limit response for glm-4.6v-flash and TheRouter parked the whole account. Cooldowns are now per model, and six GLM flash-family models have a second route through Zhipu's international service.
via Zhipu AI developer documentation