Synapse Open Synapse
Tech

Cross-request KV cache exposure in the GLM-5.3 Workers AI model, and our remediation with SGLang

Cloudflare Blog · · 16 min read

A configuration mismatch in GLM 5.3 on Workers AI triggered an SGLang bug that exposed KV cache pages across requests. Here is how we addressed the issue.

Same story elsewhere

Latest from Cloudflare Blog