---
type: "Article"
title: "DeepSeek-V4.1-Flash-1017 EXL3 1.62 bpw with vLLM on 1× DGX Spark"
description: "DeepSeek-V4.1-Flash-1017 EXL3 1.62 bpw with vLLM on 1× DGX Spark: recipe by vcruz305 with speed, condition and source for every number, weights and license."
resource: "https://www.contextstudios.ai/local-ai/vcruz305--dsv41-flash-1017-exl3-1.62-1spark-vllm"
language: "en"
generated:
  by: "process:contextstudios-md/1"
  at: "2026-11-11T03:30:12.034Z"
status: "stable"
---

# DeepSeek-V4.1-Flash-1017 EXL3 1.62 bpw with vLLM on 1× DGX Spark

DeepSeek-V4.1-Flash-1017 EXL3 1.62 bpw with vLLM on 1× DGX Spark: ~20.1 tok/s according to github.com (dataset as of Nov 3, 2026).

- Creator: [vcruz305](https://www.contextstudios.ai/local-ai/creators/vcruz305.md)
- Engine: vLLM (spark-ds41f r6w, Basis-Image dsv41-1017-ds41f:r6w) + vllm-exl3 Fork 30e8b94 + exllamav3 Fork 909ad85 (tile-K 16, W2) + Shim, DSpark k=5
- Quantization: EXL3 SAGE MixedK 1.62bpw (routed K1–K4), FP8 block-32 sonst, EXL3-K8-Head, Engram von NVMe, Cold-Expert-Alias-Datei
- Model family: DeepSeek-V4.1-Flash-1017
- Hardware: dgx-spark x1
- Context: 524288

## All numbers

- ~20.1 tok/s (Everyday; Everyday, 1 request, chat prompt (1200-token replies), sampling 0.6/0.95, with DSpark (per recipe), range 17.7–22.5 tok/s) — [source](https://github.com/vcruz305/DSV4.1-Flash-1017-EXL3-Spark-recipe#sampling-defaults)
- decode_tps: 20.1 (1 request) — [source](https://github.com/vcruz305/DSV4.1-Flash-1017-EXL3-Spark-recipe#sampling-defaults)
- decode_tps: 35.1 (1 request) — [source](https://github.com/vcruz305/DSV4.1-Flash-1017-EXL3-Spark-recipe#measured-long-context)
- prefill_tps: 720 (1 request) — [source](https://github.com/vcruz305/DSV4.1-Flash-1017-EXL3-Spark-recipe#measured-prefill)

## Weights

- [vcruz305/DSV4.1-Flash-1017-SAGE-EXL3-1.62bpw](https://huggingface.co/vcruz305/DSV4.1-Flash-1017-SAGE-EXL3-1.62bpw) (main, mit)
- [deepseek-ai/DeepSeek-V4.1-Flash-1017](https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017) (base, mit)

[https://github.com/vcruz305/DSV4.1-Flash-1017-EXL3-Spark-recipe](https://github.com/vcruz305/DSV4.1-Flash-1017-EXL3-Spark-recipe)

## Related

- [Local AI on your hardware.](https://www.contextstudios.ai/local-ai.md)