GLM-5.2

on 4× NVIDIA DGX Spark · vLLM · 131,072 ctx

Aug 28, 2026

Throughput

24.0 t/s gen · 475.0 tokens/s prompt processing (prefill)Prompt processing measures input prefill speed; it does not tell you the time to first token.

Quant

NVFP4

Summary

User reports MTP3 as the default and MTP4 as the peak. Setup includes a bug fix for a draft parallel config missing a DCP copy. Prefill reaches about 475 tps and decode at bs=3 reaches about 48 tps.

View source ↗