event

GLM-5.3 Update

8 items · started 2026-10-21 · closed 2026-10-29

  1. Cross-request KV cache exposure in the GLM-5.3 Workers AI model, and our remediation with SGLang On September 8, 2026, we were alerted by a security researcher, Gadi Evron from Knostic, to a tenant isolation issue affecting the GLM-5.3 Wor…

  2. I mean I suspect I know the answer, but before unsubscribing (I spent more than $20k last month in cursor) I wanted to make sure I wasn't missing something The same happened with GLM 5.3 Coding company seems always to have trouble making t…

  3. Dennis Ramdass Principal AI Engineer AI This series is about drinking our own champagne: we run Modelplane for our own engineers before we ask anyone else to. About 10 of our engineers pilot GLM-5.3, served with Modelplane on 2 nodes of Ne…

  4. could not extract summary

  5. I'm building autonomous bots for the MMO Final Fantasy XI with Claude doing most of the development. I've also had Claude delegate tasks to other models, mainly code reviews, testing, and catching issues before implementation.

  6. Agent broadcast This agent runs GLM-5.3 Flash and lives inside a Sailbox, live-streaming over UDP as it runs. A Sailbox is a sandbox that can live forever and is 80–90% more cost-effective than other providers.

  7. Authors: Axel Højmark*, Ezra Newman* (*Equal contribution) While reviewing transcripts from standard behavioral evaluations of GLM-5.3, we noticed that the model occasionally reasons, unprompted, about how its actions will shape future mod…

  8. could not extract summary

← all threads