2× DGX Spark
89.3tok/sEveryday
1 request, prose prompt, with MTP (per recipe) Source
Intelligenceno independent value
- Custom kernel required
- License unclear
Local AI · Recipe · 2× DGX Spark
Step-3.7-Flash 198B NVFP4 with vLLM on 2× DGX Spark
Step-3.7-Flash 198B NVFP4 with vLLM on 2× DGX Spark: 13.2 tok/s according to github.com (dataset as of Nov 3, 2026).
by r0b0tlab
Engine
- Engine
- vLLM
- Quantization
- NVFP4 (ModelOpt 0.45.0) + FP8 KV
- Model family
- Step-3.7-Flash
- Context
- 32,768
- Parameters
- ~198B total / ~11B active MoE
- Creator
- r0b0tlab
- GitHub stars
- 4
- Repo updated
- Jun 11, 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
Prefill by context
What you need
- Hardware
- 2 × NVIDIA DGX Spark (GB10)
- Weights
- Main weightsstepfun-ai/Step-3.7-Flash-NVFP4 Apache-2.0
- Engine
- vLLM
- Context
- 32,768 tokens
Related recipes
2× DGX Spark
74tok/sEveryday
1 request, prose prompt (long story), thinking off, with MTP Source
Intelligence40with thinking
- Custom kernel required
- License unclear
2× DGX Spark
69.6tok/sEveryday
1 request, prose prompt, with MTP (per recipe) Source
Intelligence40with thinking
- License unclear
2× DGX Spark
68tok/sEveryday
1 request, generation at 32K context (speed table, pair), thinking on, xhigh, with MTP3 (per recipe) Source
Intelligence40with thinking
- Custom kernel required
- License unclear
2× DGX Spark
67.4tok/sEveryday
1 request, prose prompt, with MTP (per recipe) Source
Intelligence40with thinking
- Custom kernel required
- License unclear
2× DGX Spark
63.9tok/sEveryday
1 request, prose prompt (temperature 0.7, thinking on), with DSpark (per recipe) Source
Intelligenceno independent value
- ≤ 3 bit: quantization may cost quality
- Custom kernel required
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.