FR
Réserver un appel

Local AI · Recipe · 2× DGX Spark

GLM-5.3-Flash NVFP4 with vLLM on 2× DGX Spark (128k context, with DFlash2)

GLM-5.3-Flash NVFP4 with vLLM on 2× DGX Spark (128k context, with DFlash2): 20.5 tok/s according to github.com (dataset as of Sep 29, 2026).

by Weschera

20.5tok/sEveryday

1 request, prose prompt, no speculative decoding

Source: github.com
Intelligence (original model) · Artificial Analysisno independent value
Instructions in the README Repo

Engine

Engine
vLLM
Quantization
NVFP4 (Marlin W4A16); FP8 E4M3 KV (6 GiB-Pin)
Model family
GLM-5.3-Flash
Context
131,072
Creator
Weschera
GitHub stars
0
Repo updated
Sep 10, 2026

Measurements

Every sourced value of this recipe, each with its condition and source. Bars relative to the largest value in the group.

Decode · 1 request

  • Everyday (decode)20.5 tok/sMedian aus 4 sustained Requests je Arm (2 EN-Tutorials, 2 KO-Essays, c=1); DFlash2 k=7 SELECTED (Code-Contract bestanden, 2 strikte Format-Fails); Baseline ohne Spec 15.18 (+35 %) Source 
  • Peak (decode)20.5 tok/sMedian aus 4 sustained Requests je Arm (2 EN-Tutorials, 2 KO-Essays, c=1); DFlash2 k=7 SELECTED (Code-Contract bestanden, 2 strikte Format-Fails); Baseline ohne Spec 15.18 (+35 %) Source 

What you need

Hardware
2 × NVIDIA DGX Spark (GB10)
Weights
Engine
vLLM
Context
131,072 tokens

Sources

Related recipes

2× DGX Spark

GLM-5.3-Flash EXL3 with TensorFold on 2× DGX Spark (1M context, 4 streams)

by MiaAI-Lab

97.6tok/sEveryday

HumanEval 97.6% / GSM8K 98.0% with FP8 KV and 4-bit dense Source 

Intelligence42with thinking
  • ≤ 3 bit: quantization may cost quality
  • Custom kernel required
  • Non-commercial
  • License unclear

EXL3/TR3 4bpw routed experts + 4-bit dense, BF16 elsewhere, FP8 KVTensorFold v0.5.0 (52 patches: DFlash2+copy drafts, 4-bit dense, FP8 KV, RoCE one-shot all-gather, shared KV pool, vision, tool calling)Weights Repo Updated Sep 30, 2026

Details

← Back to overview

Local AI in your company?

In a workshop we work out which models and which hardware fit your tasks, and build the first agent on your infrastructure.

See the AI workshopBack to overview