---
type: "Article"
title: "Spark-X2.5-4B BF16 with SGLang on 1× DGX Spark"
description: "Spark-X2.5-4B BF16 with SGLang on 1× DGX Spark: recipe by Weschera with speed, condition and source for every number, weights and license."
resource: "https://www.contextstudios.ai/local-ai/weschera--spark-x25-4b-bf16-sglang-1x-dgx-spark"
language: "en"
generated:
  by: "process:contextstudios-md/1"
  at: "2026-10-02T21:59:02.078Z"
status: "stable"
---

# Spark-X2.5-4B BF16 with SGLang on 1× DGX Spark

Spark-X2.5-4B BF16 with SGLang on 1× DGX Spark: 21.8 tok/s according to github.com (dataset as of Sep 29, 2026).

- Creator: [Weschera](https://www.contextstudios.ai/local-ai/creators/weschera.md)
- Engine: SGLang
- Quantization: BF16 (unquantisiert)
- Model family: Spark-X2.5-4B
- Hardware: dgx-spark x1
- Context: 131072

## All numbers

- 21.8 tok/s (Everyday; Everyday, 1 request, prose prompt, no speculative decoding) — [source](https://github.com/Weschera/spark-x25-4b-dgx-recipe/blob/main/README.md)
- decode_tps: 21.83 (1 request) — [source](https://github.com/Weschera/spark-x25-4b-dgx-recipe/blob/main/README.md)
- decode_tps: 21.83 (1 request) — [source](https://github.com/Weschera/spark-x25-4b-dgx-recipe/blob/main/README.md)

## Weights

- [XHToken/Spark-X2.5-4B](https://huggingface.co/XHToken/Spark-X2.5-4B) (main, apache-2.0)

[https://github.com/Weschera/spark-x25-4b-dgx-recipe](https://github.com/Weschera/spark-x25-4b-dgx-recipe)

## Related

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