4× RTX PRO 6000
235tok/sEveryday
1 request, prose prompt, context 19, with DSpark (per recipe) Source
Intelligence34with thinking
- Custom kernel required
Local AI · Recipe · 1× DGX Spark
DeepSeek-V4-Flash 284B EXL3 2.54 bpw with ExLlamaV3 on 1× DGX Spark
DeepSeek-V4-Flash 284B EXL3 2.54 bpw with ExLlamaV3 on 1× DGX Spark: 30.2 tok/s according to github.com (dataset as of Nov 3, 2026).
by 0xBakeer
Intelligence (original model) · Artificial Analysis34with thinking
≤ 3 bit: quantization may cost quality
Engine
- Engine
- ExLlamaV3 (aarch64-Port)
- Quantization
- EXL3 2.54 bpw (92 GB)
- Model family
- DeepSeek-V4-Flash
- Context
- 393,216
- Parameters
- 284B MoE
- Creator
- 0xBakeer
- GitHub stars
- 3
- Repo updated
- Sep 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
Prefill by context
Time to first token
Shorter is better.
What you need
- Hardware
- 1 × NVIDIA DGX Spark (GB10)
- Weights
- Main weightsanoane/DeepSeek-V4-Flash-0731-exl3-2.54bpw MIT
- Engine
- ExLlamaV3 (aarch64-Port)
- Context
- 393,216 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.
Sources
Related recipes
4× DGX Spark
46tok/sEveryday
1 request, prose prompt, with DSpark Source
Intelligenceno independent value
4× DGX Spark
42tok/sEveryday
1 request, prose prompt, with DSpark (per recipe) Source
Intelligenceno independent value
- Custom kernel required
3× DGX Spark
41.9tok/sEveryday
1 request, prose prompt, with DSpark (per recipe) Source
Intelligenceno independent value
- Custom kernel required
2× DGX Spark
33.2tok/sEveryday
1 request, prose prompt, with DSpark (per recipe) Source
Intelligenceno independent value
- Custom kernel required
1× Mac
28.5tok/sEveryday
1 request, prose prompt, context 32k, with DSpark (per recipe) Source
Intelligenceno independent value
- ≤ 3 bit: quantization may cost quality
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.