Matheus Nogueira·5d agoOne Recipe, Many Distributions — and How Not to Fool Yourself: GLMs, Cross-Validation and the…The GLM recipe goes beyond classification. Then: how to estimate test error with cross-validation, and uncertainty with the bootstrap.
rhythm·Oct 30GLM-5.3-Flash and the Work That Comes After an AI AnswerThree tasks showed where this multimodal model was useful — and where a convincing response still fell short of a finished result.
rhythm·Oct 22GLM-5 vs Gemini 3.8 Flash: Two Failures a Pass Rate Cannot ShowBoth calls returned 200 and raised no error. The difference is only visible in the transcript.
UplinkSwitch·Oct 17GLM API cost per request, October 2026: a chat turn and a long documentWhat one GLM API request costs on GLM 5.3, 5.2 and 5.3 Flash: a chat turn and a long document, worked at Z.ai’s list price and ours.
JV·Oct 12How to run GLM-5.3-Flash on a Mac Studio with oMLX: install, serve, DFlash and a coding agentTL;DR
Denton Liu·Oct 11How to Choose Between GLM-4.7 and Other Coding ModelsFor most teams, GLM-4.7 wins on cost-per-resolved-task rather than raw benchmark scores; it is the model you route routine coding work to…
Nitin Gavhane·Oct 2GLM-5.3: The Open-Weight AI Model That’s Redefining Cybersecurity — And How to Run ItIn August 2026, Zhipu AI (Z.ai) released GLM-5.3, a 753-billion-parameter Mixture-of-Experts model that has sent shockwaves through both…
AI Engineering·Sep 17GLM Built Its Own Inference Infrastructure: 2 Weeks, 100K Accelerators, 3x Higher Throughput
