GLM-5.1

Z.ai · 753.9B (active 40B) · MoE · ctx 198K · MIT · HuggingFace

GLM-5.1 (Z.ai) is a 753.9B parameters MoE (40B active per token) open-weight model tracked on AI Hashrate. At 8K context we estimate roughly 423.65 GB VRAM for Q4 (weights 414.65 GB + KV 8.0 GB + 1 GB overhead) and 1516.8 GB for FP16. About 0 GPUs/accelerators in our catalog fully fit this model at Q4 under that context assumption. Fastest Q4 configs: B200 192GB SXM ≈ 74.7 tok/s (estimated); MI300X 192GB ≈ 49.5 tok/s (estimated); H200 141GB SXM5 ≈ 24.2 tok/s (estimated). Use the table below for VRAM fit and tok/s per dollar (list MSRP). Estimates follow memory-bandwidth math; measured rows override when present. See methodology for details. Methodology.

7 hardware configs — sorted by tok/s. Default fit context 8K.