---

license: apache-2.0

base_model:

• Qwen/Qwen3.5-9B
• XiaomiMiMo/MiMo-V2.6-Distill-Qwen-9B
• Jackrong/Qwen3.5-9B-Claude-4.6-Opus-Reasoning-Distilled-v2
• DavidAU/Qwen3.5-9B-Claude-4.6-HighIQ-THINKING-HERETIC-UNCENSORED

base_model_relation: merge

library_name: gguf

pipeline_tag: text-generation

tags: [gguf, merge, qwen3.5, winery, llama.cpp, model-soup]

---

<p align="center"><a href="https://huggingface.co/WineryLabs"><img src="https://winery-api-zandy.zocomputer.io/card/m/9b.svg" alt="Winery Qwen3.5-9B-GrandCru" width="100%"/></a></p>

🍷 Winery Qwen3.5 9B · Grand Cru

A score-weighted model soup of the three strongest Qwen3.5-9B fine-tunes out of the 10 we benchmarked, blended with the Winery fusion compiler. It's 9.5 GB at Q8_0.

Recipe

Linear merge. Each donor's weight is its own benchmark average:

| Donor | Weight (its avg) |

|---|---|

| XiaomiMiMo/MiMo-V2.6-Distill-Qwen-9B | 84.9 |

| Jackrong/Qwen3.5-9B-Claude-4.6-Opus-Reasoning-Distilled-v2 | 83.4 |

| DavidAU/Qwen3.5-9B-Claude-4.6-HighIQ-THINKING-HERETIC-UNCENSORED | 83.2 |

name Winery 9B Grand Cru
out q8_0
default linear 84.9,83.4,83.2

Scores (local quick bench, 0-shot chat, thinking off)

MMLU 200 / ARC-Challenge 150 / GSM8K 60 questions, run through llama-server.

| Model | MMLU | ARC-C | GSM8K | Avg |

|---|---|---|---|---|

| Winery 9B Grand Cru (this) | 70.0 | 96.7 | 90.0 | 85.6 |

| Best donor (MiMo-V2.6 Distill) | 69.5 | 95.3 | 90.0 | 84.9 |

| Qwen/Qwen3.5-9B (stock) | 69.5 | 94.7 | 66.7 | 77.0 |

These are small-sample numbers from our own harness, so treat them as directional only. They aren't leaderboard results.

Use

Works in llama.cpp (recent builds with Qwen3.5 support), Jan, LM Studio and the Winery app. Tested in Jan.

llama-cli -m Winery-Qwen3.5-9B-GrandCru-Q8_0.gguf -cnv

⚠️ One donor is an uncensored/abliterated tune, so this model refuses less than stock Qwen. You're responsible for how you use it.