2× DGX Spark
35tok/sEveryday
1 request, prose prompt, with EAGLE Source
Intelligenceno independent value
- Experimental
Local AI · Recipe · 1× RTX PRO 6000
MiMo-V2.6-Flash EXL3 2.20 bpw with ExLlamaV3 on 1× RTX PRO 6000
by vcruz305
Intelligence (original model) · Artificial Analysisno independent value
≤ 3 bit: quantization may cost quality
Engine
- Engine
- ExLlamaV3 (native /v1 server) v1.5.1.post1 (Fork)
- Quantization
- EXL3 2.20bpw (86.94 GB)
- Model family
- MiMo-V2.6-Flash
- Context
- 65,536
- Parameters
- 48 Layer, 256 routed Experten top-8
- Creator
- vcruz305
- GitHub stars
- 0
- Repo updated
- Sep 25, 2026
Measurements
Every sourced value of this recipe, each with its condition and source. Bars relative to the largest value in the group.
No sourced measurements for this recipe.
What you need
- Hardware
- 1 × NVIDIA RTX PRO 6000 Blackwell
- Weights
- Main weightsvcruz305/MiMo-V2.6-Flash-RL-EXL3 MIT
- Draft modelXiaomiMiMo/MiMo-V2.6-Flash-RL MIT
- BaseXiaomiMiMo/MiMo-V2.6-Flash-RL MIT
- Engine
- ExLlamaV3 (native /v1 server) v1.5.1.post1 (Fork)
- Context
- 65,536 tokens
Notes
What matters before you rebuild it.
- ≤ 3 bit: quantization may cost qualityAt 3 bit and below the model may answer noticeably worse than the original. The intelligence number refers to the original.
- Custom kernel required
Sources
Related recipes
2× DGX Spark
53.3tok/sMixed
1 request, mixed prompt set, with DFlash (per recipe) Source
Intelligenceno independent value
- Custom kernel required
2× DGX Spark
no documented speed
Intelligenceno independent value
- Custom kernel required
1× RTX PRO 6000
207tok/sMixed
1 request, mixed prompt set, with FR-Spec + NEXTN (per recipe) Source
Intelligence40with thinking
- Custom kernel required
- License unclear
1× RTX PRO 6000
287tok/sPeak
1 request, synthetic prompt Source
Intelligence9with thinking
1× RTX PRO 6000
281tok/sPeak
1 request, synthetic prompt Source
Intelligenceno independent value
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.