---
type: "Article"
title: "MiniCPM-V 4.7 35B-A3B EXL3 3.74 bpw with ExLlamaV3 on 1× DGX Spark"
description: "MiniCPM-V 4.7 35B-A3B EXL3 3.74 bpw with ExLlamaV3 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--minicpm-v4.7-35b-exl3-sage-3.74-1spark-tabby"
language: "en"
generated:
  by: "process:contextstudios-md/1"
  at: "2026-11-11T03:30:20.822Z"
status: "stable"
---

# MiniCPM-V 4.7 35B-A3B EXL3 3.74 bpw with ExLlamaV3 on 1× DGX Spark

MiniCPM-V 4.7 35B-A3B EXL3 3.74 bpw with ExLlamaV3 on 1× DGX Spark: 60.3 tok/s according to github.com (dataset as of Nov 3, 2026).

- Creator: [vcruz305](https://www.contextstudios.ai/local-ai/creators/vcruz305.md)
- Engine: ExLlamaV3 Fork vcruz305 (master, gemessen dc778646 / 475461e9) + TabbyAPI Fork vcruz305
- Quantization: EXL3 SAGE 3.74bpw (Kopf/Vision 8-bit), FP16-KV (8K) bzw. Q4-KV (262K)
- Model family: MiniCPM-V-4.7-35B-A3B
- Hardware: dgx-spark x1
- Context: 262144

## All numbers

- 60.3 tok/s (Everyday; Everyday, 1 request, prose prompt (essay), context 2k, no drafter (per recipe)) — [source](https://github.com/vcruz305/MiniCPM-V-4.7-35B-A3B-EXL3-DGX-Spark-recipe#dspark-exact-speculative-decoding)
- decode_tps: 60.3 (1 request) — [source](https://github.com/vcruz305/MiniCPM-V-4.7-35B-A3B-EXL3-DGX-Spark-recipe#dspark-exact-speculative-decoding)
- decode_tps: 69.48 (1 request) — [source](https://github.com/vcruz305/MiniCPM-V-4.7-35B-A3B-EXL3-DGX-Spark-recipe#dspark-exact-speculative-decoding)
- prefill_tps: 1056 (1 request) — [source](https://github.com/vcruz305/MiniCPM-V-4.7-35B-A3B-EXL3-DGX-Spark-recipe#full-context-results-one-dgx-spark)
- ttft_ms: 948 (1 request) — [source](https://github.com/vcruz305/MiniCPM-V-4.7-35B-A3B-EXL3-DGX-Spark-recipe#dspark-exact-speculative-decoding)

## Weights

- [vcruz305/MiniCPM-V-4.7-35B-A3B-EXL3-SAGE-3.74bpw](https://huggingface.co/vcruz305/MiniCPM-V-4.7-35B-A3B-EXL3-SAGE-3.74bpw) (main, apache-2.0)
- [openbmb/MiniCPM-V-4.7-35B-A3B](https://huggingface.co/openbmb/MiniCPM-V-4.7-35B-A3B) (base, apache-2.0)

[https://github.com/vcruz305/MiniCPM-V-4.7-35B-A3B-EXL3-DGX-Spark-recipe](https://github.com/vcruz305/MiniCPM-V-4.7-35B-A3B-EXL3-DGX-Spark-recipe)

## Related

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