Elevated GLM-5.3-Flash Errors Due to Incorrect Model Routing
Resolved · Major · Nov 6, 1:30 AM · 24h
z-ai/glm-5.3-flash-abliterated · z-ai/glm-5.3-flash · Platform API
Resolved
During the initial deployment of GLM-5.3-Flash Abliterated, a shared model flag caused the fleet manager to route all GLM-5.3-Flash requests to the abliterated version, resulting in an increased error rate for users. Now resolved