4× DGX Spark
35.7tok/sEveryday
1 request, prose prompt, with DFlash2, thinking on (per recipe) Source
Intelligence45with thinking
- Custom kernel required
- Non-commercial
Local AI · Recipe · 8× DGX Spark
GLM-5.3 NVFP4 with SparkRing on 8× DGX Spark
GLM-5.3 NVFP4 with SparkRing on 8× DGX Spark: 46 tok/s according to github.com (dataset as of Nov 3, 2026).
46tok/sEveryday
1 request, generation at 32K context (speed table, ring of eight, TP8/DCP4), thinking always on (measured before the profile default high, i.e. template default max), with MTP2 (per recipe); experimental profile
Source: github.comEngine
- Engine
- vLLM (local-inference-lab Fork) via SparkRing (SIRCL-Collectives, switchless RDMA ueber ConnectX-7), Image Release 2026.10.2
- Quantization
- NVFP4-Experten/BF16 (Local Inference Lab; Dense-Linears beim Laden MXFP8)
- Model family
- GLM-5.3
- Context
- 1,048,576
- Parameters
- 744B MoE (GLM-5.3 laut Profil-README); 464,9 GB Safetensors (432,9 GiB)
- Creator
- fujitsupolycom
- GitHub stars
- 102
- Repo updated
- Nov 2, 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
Decode · parallel requests
Sums over all streams are marked; per request it is correspondingly less.
Prefill by context
What you need
- Hardware
- 8 × NVIDIA DGX Spark (GB10)
- Weights
- Main weightslocal-inference-lab/GLM-5.3-NVFP4 MIT
- Engine
- vLLM (local-inference-lab Fork) via SparkRing (SIRCL-Collectives, switchless RDMA ueber ConnectX-7), Image Release 2026.10.2
- Context
- 1,048,576 tokens
Notes
What matters before you rebuild it.
- Custom kernel required
- Experimental
Sources
Related recipes
3× DGX Spark
31.3tok/sEveryday
1 request, prose prompt, with DSpark (per recipe) Source
Intelligence45with thinking
- Custom kernel required
- License unclear
1× Mac
6.14tok/sMixed
1 request, mixed prompt set, context 1k, no speculative decoding Source
Intelligence45with thinkingOriginal model's value. Heavily compressed (~2.1 bit), intelligence is likely much lower.
- ≤ 3 bit: quantization may cost quality
- Streams from SSD
- License unclear
4× DGX Spark
53.3tok/sPeak
1 request, structured output end-to-end (count-to-300); DCP4 lane 2: NVFP4 KV Source
Intelligence45with thinking
- Custom kernel required
- License unclear
4× DGX Spark
~36tok/sPeakapproximate · method not stated
1 request, prompt type not stated, range 33–39 tok/s, with MTP-3 (per recipe) Source
Intelligence45with thinking
- ≤ 3 bit: quantization may cost quality
- Custom kernel required
- License unclear
8× DGX Spark
74tok/sEveryday
1 request, generation at 32K context (speed table, ring of eight), thinking on, high, with DSpark (per recipe); experimental profile Source
Intelligence40with thinkingno thinking 25
- Custom kernel required
- Experimental
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.