1× DGX Spark
79.7tok/sMixed
1 request, mixed prompt set, with MTP (per recipe) Source
Intelligence18with thinkingno thinking 15
Local AI · Recipe · 1× DGX Spark
Qwen3.6-35B-A3B NVFP4 with vLLM on 1× DGX Spark (256k context)
Qwen3.6-35B-A3B NVFP4 with vLLM on 1× DGX Spark (256k context): 93.1 tok/s according to github.com (dataset as of Nov 3, 2026).
by r0b0tlab
Intelligence (original model) · Artificial Analysis18with thinkingno thinking 15
Engine
- Engine
- vLLM 0.25.0 (702f4814)
- Quantization
- NVFP4 + FP8 KV
- Model family
- Qwen3.6-35B-A3B
- Context
- 262,144
- Parameters
- 35B total / 3B active MoE
- Creator
- r0b0tlab
- GitHub stars
- 7
- Repo updated
- Jul 15, 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
Time to first token
Shorter is better.
What you need
- Hardware
- 1 × NVIDIA DGX Spark (GB10)
- Weights
- Main weightsnvidia/Qwen3.6-35B-A3B-NVFP4 Apache-2.0
- Engine
- vLLM 0.25.0 (702f4814)
- Context
- 262,144 tokens
Related recipes
1× DGX Spark
57.5tok/sMixed
1 request, mixed prompt set, with MTP Source
Intelligence18with thinkingno thinking 15
1× DGX Spark
45.4tok/sMixed
1 request, mixed prompt set, no speculative decoding Source
Intelligence18with thinkingno thinking 15
1× Strix Halo
70.2tok/sPeak
1 request, synthetic prompt Source
Intelligence18with thinkingno thinking 15
1× DGX Spark
71.8tok/sEveryday
1 request, prose prompt (sampled, thinking off), with LiLiCorr (per recipe) Source
Intelligenceno independent value
- ≤ 3 bit: quantization may cost quality
- Custom kernel required
- License unclear
1× DGX Spark
60.3tok/sEveryday
1 request, prose prompt (essay), context 2k, no drafter (per recipe) Source
Intelligenceno independent value
- 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.