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
DeepSeek-V4.1-Flash 763B EXL3 3.5 bpw with ExLlamaV3 on 4× DGX Spark
by tonyd2wild
Intelligence (original model) · Artificial Analysis40with thinkingno thinking 25
Engine
- Engine
- vLLM + ExLlamaV3 (Default), SGLang experimentell
- Quantization
- EXL3 3.5bpw routed experts + FP8 Rest
- Model family
- DeepSeek-V4.1-Flash
- Context
- 500,000
- Parameters
- ~763B
- Creator
- tonyd2wild
- GitHub stars
- 86
- Repo updated
- Sep 24, 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
- 4 × NVIDIA DGX Spark (GB10)
- Weights
- Main weightsbot-lab-21/DeepSeek-V4.1-Flash-EXL3-3.5bpw-Pollard MIT
- Basedeepseek-ai/DeepSeek-V4.1-Flash MIT
- Engine
- vLLM + ExLlamaV3 (Default), SGLang experimentell
- Context
- 500,000 tokens
Notes
What matters before you rebuild it.
- Experimental
- Custom kernel required
Sources
Related recipes
4× RTX PRO 6000
73.1tok/sEveryday
1 request, prose prompt, context 32788, with DSPARK Source
Intelligence40with thinkingno thinking 25
4× DGX Spark
59.1tok/sEveryday
1 request, prose prompt, with DSpark (per recipe) Source
Intelligence40with thinkingno thinking 25
3× DGX Spark
51tok/sEveryday
1 request, prose prompt, with MTP/EAGLE (spekulativ, ~59 ms/step) (per recipe) Source
Intelligence40with thinkingno thinking 25
- Custom kernel required
2× DGX Spark
31.6tok/sEveryday
1 request, prose prompt, with DSpark (per recipe) Source
Intelligence40with thinkingno thinking 25
- ≤ 3 bit: quantization may cost quality
- Custom kernel required
2× DGX Spark
31.6tok/sEveryday
1 request, prose prompt, with DSpark (per recipe) Source
Intelligence40with thinkingno thinking 25
- ≤ 3 bit: quantization may cost quality
- Experimental
- 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.