---
type: "Article"
title: "MiniCPM-V-4.7-35B-A3B EXL3 6.08 bpw with ExLlamaV3 on 1× DGX Spark (text, 262k context)"
description: "MiniCPM-V-4.7-35B-A3B EXL3 6.08 bpw with ExLlamaV3 on 1× DGX Spark (text, 262k context): recipe by r0b0tlab with speed, condition and source for every number, weights and license."
resource: "https://www.contextstudios.ai/local-ai/r0b0tlab--minicpm-v4.7-35b-a3b-exl3-6.08bpw-exllamav3-1x-spark"
language: "en"
generated:
  by: "process:contextstudios-md/1"
  at: "2026-11-11T03:30:27.943Z"
status: "stable"
---

# MiniCPM-V-4.7-35B-A3B EXL3 6.08 bpw with ExLlamaV3 on 1× DGX Spark (text, 262k context)

MiniCPM-V-4.7-35B-A3B EXL3 6.08 bpw with ExLlamaV3 on 1× DGX Spark (text, 262k context): 59.4 tok/s according to github.com (dataset as of Nov 3, 2026).

- Creator: [r0b0tlab](https://www.contextstudios.ai/local-ai/creators/r0b0tlab.md)
- Engine: ExLlamaV3 1.5.0 (Fork r0b0tlab/exllamav3-r0b0t, Branch minicpm-v4.7-dspark @ 5ddae7d12448), torch 2.14.0+cu130
- Quantization: EXL3 6.08 bpw (28,3 GB), FP16 KV; nur Textpfad (Vision-Tower wird von ExLlamaV3 nicht ausgeführt)
- Model family: MiniCPM-V-4.7-35B-A3B
- Hardware: dgx-spark x1
- Context: 262144

## All numbers

- 59.4 tok/s (Everyday; Everyday, 1 request, chat prompt (20 prompts, thinking off), no drafting (default profile)) — [source](https://github.com/r0b0tlab/minicpm-v47-exl3-exllamav3-gb10#results)
- decode_tps: 59.4 (1 request) — [source](https://github.com/r0b0tlab/minicpm-v47-exl3-exllamav3-gb10#results)
- decode_tps: 59.9 (1 request) — [source](https://github.com/r0b0tlab/minicpm-v47-exl3-exllamav3-gb10/blob/main/docs/RESULTS.md#leg-d-plain-vs-drafted-decode)
- decode_tps: 59.8 (1 request) — [source](https://github.com/r0b0tlab/minicpm-v47-exl3-exllamav3-gb10#results)
- prefill_tps: 2889 (1 request) — [source](https://github.com/r0b0tlab/minicpm-v47-exl3-exllamav3-gb10#results)
- ttft_ms: 11300 (1 request) — [source](https://github.com/r0b0tlab/minicpm-v47-exl3-exllamav3-gb10#results)

## Weights

- [r0b0tlab/MiniCPM-V-4.7-35B-A3B-EXL3-6.08bpw](https://huggingface.co/r0b0tlab/MiniCPM-V-4.7-35B-A3B-EXL3-6.08bpw) (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/r0b0tlab/minicpm-v47-exl3-exllamav3-gb10](https://github.com/r0b0tlab/minicpm-v47-exl3-exllamav3-gb10)

## Related

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