Local AI · Recipe · 2× DGX Spark

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

GLM-5.3-Flash NVFP4 with vLLM on 2× DGX Spark (128k context, with MTP): 23 tok/s according to github.com (dataset as of Nov 3, 2026).

by Weschera

23tok/sEveryday

1 request, prose prompt, with MTP4

Source: github.com
Intelligence (original model) · Artificial Analysisno independent value

Engine

Engine
vLLM
Quantization
NVFP4 (vcruz305-Quant, BF16-MTP-Head, 49 Shards; quantization_config.ignore für Layer 45), KV fp8_e4m3
Model family
GLM-5.3-Flash
Context
131,072
Creator
Weschera
GitHub stars
0
Repo updated
Aug 28, 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)23 tok/sBereich 22–24 tok/s, MTP4, thinking off, short/medium ctx → approx (Bereichsangabe) Source 
  • Peak (decode)36.5 tok/sStructured output/code, MTP4 Source 

What you need

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

Notes

What matters before you rebuild it.

  • Experimental
  • Custom kernel required

Sources

Related recipes

4× DGX Spark
68tok/sEveryday

1 request, generation at 32K context (speed table; measured only as a path of four Sparks inside a ring of eight, not as a ring of four), thinking always on (measured before the profile default high, i.e. template default max), with MTP3 (per recipe) Source 

Intelligence42with thinking
  • Custom kernel required
  • License unclear

NVFP4/MXFP8 CSF QAD (Local Inference Lab)vLLM (local-inference-lab Fork) via SparkRing (SIRCL-Collectives, switchless RDMA ueber ConnectX-7), Image Release 2026.10.2Weights Repo Updated Nov 2, 2026

Details
2× DGX Spark
60.4tok/sEveryday

Prose decode, 1 request; quality (FP8 KV, 4-bit dense): GSM8K 98.8% / HumanEval 95.7% / HumanEval+ & MBPP+ 86.5% 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 v0.6.0 + Recipe v1.10 (README: 96 Patches; gepinntes Image v0.6.0-e66c657a433d mit 107 Patches: DFlash2+Copy-Drafts, 4-bit dense, FP8-KV, RoCE-One-Shot-All-Gather, Shared KV-Pool, optionaler Spill-Tier, Vision, Tool-Calling, Anthropic /v1/messages; PARALLEL=8 opt-in)Weights Repo Updated Oct 25, 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.