4× DGX Spark
87.7tok/sEveryday
1 request, prose prompt, with DSpark (per recipe) Source
Intelligence40with thinkingno thinking 25
Local AI · Recipe · 4× DGX Spark
MiMo-V2.6-Pro-MOPD ARVQ/NVFP4 hybrid with vLLM on 4× DGX Spark (1M context)
MiMo-V2.6-Pro-MOPD ARVQ/NVFP4 hybrid with vLLM on 4× DGX Spark (1M context): 24.2 tok/s according to github.com (dataset as of Nov 3, 2026).
by drowzeys
Intelligence (original model) · Artificial Analysisno independent value
Engine
- Engine
- vLLM-Fork jarrelscy/vllm-mimo-v26-arvq-sm120 (Image ghcr.io/drowzeys/mimo-v26-pro-arvq-spark:63430f7-sm121-v4)
- Quantization
- Jarrelscy ARVQ/NVFP4-Hybrid (hybrid-21: 21% NVFP4-Hot-Experten), BF16 KV
- Model family
- MiMo-V2.6-Pro
- Context
- 1,048,576
- Creator
- drowzeys
- GitHub stars
- 8
- Repo updated
- Sep 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
Decode · parallel requests
Sums over all streams are marked; per request it is correspondingly less.
Prefill by context
What you need
- Hardware
- 4 × NVIDIA DGX Spark (GB10)
- Weights
- Main weightsdrowzeys/keys-MiMo-V2.6-Pro-MOPD-Jarrelscy-ARVQ-hybrid21 MIT
- BaseXiaomiMiMo/MiMo-V2.6-Pro-MOPD MIT
- otherjarrelscy/MiMo-V2.6-Pro-RL-ARVQ-hybrid MIT
- Engine
- vLLM-Fork jarrelscy/vllm-mimo-v26-arvq-sm120 (Image ghcr.io/drowzeys/mimo-v26-pro-arvq-spark:63430f7-sm121-v4)
- Context
- 1,048,576 tokens
Notes
What matters before you rebuild it.
- Fork
- Multiple nodes
- Custom kernel required
- tool-calling
Sources
Related recipes
4× DGX Spark
78tok/sEveryday
1 request, prose prompt, with DSpark (per recipe) Source
Intelligenceno independent value
- Custom kernel required
4× DGX Spark
73tok/sEveryday
1 request, decode at 32K context (speed table, ring of four), thinking on, xhigh Source
Intelligence40with thinking
- Custom kernel required
4× DGX Spark
71tok/sEveryday
1 request, decode at 32K context (speed table, ring of four), thinking on, high Source
Intelligence40with thinkingno thinking 25
- Custom kernel required
4× DGX Spark
70.9tok/sEveryday
1 request, prose prompt (greedy, thinking off), with DSpark (per recipe) Source
Intelligenceno independent value
- ≤ 3 bit: quantization may cost quality
- Custom kernel required
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
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.