GLM 5.3 Flash2× RTX PRO 6000 · EXL3 · TensorFold 0.6.5Decode · tok/s238Prefill—tok/100W39.72x RTX PRO 6000 (300 W per GPU): 238 tok/s on prose, single streamKurcide print, quoted by Mia. Rank is the quoted 238 prose one-stream.Officialclaimed@KurcideAevonix recipeWatchCopy linkOpen pair