---
title: Qwen3-18B-A3B-Stranger-Thoughts-Q_8
canonical_url: "https://www.modelscope.cn/models/mlx-community/Qwen3-18B-A3B-Stranger-Thoughts-Q_8"
md_url: "https://www.modelscope.cn/models/mlx-community/Qwen3-18B-A3B-Stranger-Thoughts-Q_8.md"
repository: mlx-community/Qwen3-18B-A3B-Stranger-Thoughts-Q_8
last_updated: 2025-06-29
license: apache-2.0
pipeline_tag: text-generation
tasks:
  - text-generation
model_type:
  - qwen3_moe
architectures:
  - Qwen3MoeForCausalLM
base_model:
  - DavidAU/Qwen3-18B-A3B-Stranger-Thoughts
base_model_relation: quantized
parameters: 4.9B
tensor_type:
  - U32
  - BF16
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
language:
  - en
  - fr
  - zh
  - de
inference_backends:
  - "deploy_task text/emb"
  - "lmdeploy_turbomind 0.9.1"
  - "sglang 0.5.2"
  - "vllm 0.9.2"
downloads: 207
stars: 0
tags:
  - creative
  - "creative writing"
  - "fiction writing"
  - "plot generation"
  - "sub-plot generation"
  - "story generation"
  - "scene continue"
  - storytelling
  - "fiction story"
  - "science fiction"
  - romance
  - "all genres"
  - story
  - writing
  - "vivid prose"
  - "vivid writing"
  - moe
  - "mixture of experts"
  - "64 experts"
  - "8 active experts"
  - fiction
  - roleplaying
  - bfloat16
  - rp
  - qwen3
  - horror
  - finetune
  - thinking
  - reasoning
  - qwen3_moe
  - mlx
---

# Qwen3-18B-A3B-Stranger-Thoughts-Q_8

> Qwen3-18B-A3B-Stranger-Thoughts-Q_8 - mlx-community 在 ModelScope 开源的模型。mlx-community/Qwen3-18B-A3B-Stranger-Thoughts-Q8

mlx-community/Qwen3-18B-A3B-Stranger-Thoughts-Q_8 是 ModelScope 魔搭社区上的 4.9B 参数text-generation模型，采用 apache-2.0 许可，基于 DavidAU/Qwen3-18B-A3B-Stranger-Thoughts 构建，可用 deploy_task text/emb、lmdeploy_turbomind 0.9.1、sglang 0.5.2 部署。

- **Repository**: mlx-community/Qwen3-18B-A3B-Stranger-Thoughts-Q_8
- **License**: apache-2.0
- **Tasks**: text-generation
- **Parameters**: 4.9B
- **Base model**: DavidAU/Qwen3-18B-A3B-Stranger-Thoughts
- **Inference backends**: deploy_task text/emb, lmdeploy_turbomind 0.9.1, sglang 0.5.2, vllm 0.9.2
- **Tags**: creative, creative writing, fiction writing, plot generation, sub-plot generation, story generation, scene continue, storytelling, fiction story, science fiction, romance, all genres, story, writing, vivid prose, vivid writing, moe, mixture of experts, 64 experts, 8 active experts, fiction, roleplaying, bfloat16, rp, qwen3, horror, finetune, thinking, reasoning, qwen3_moe, mlx
- **Downloads**: 207
- **Stars**: 0
- **Last updated**: 2025-06-29

Source: https://www.modelscope.cn/models/mlx-community/Qwen3-18B-A3B-Stranger-Thoughts-Q_8

---

# mlx-community/Qwen3-18B-A3B-Stranger-Thoughts-Q_8

This model [mlx-community/Qwen3-18B-A3B-Stranger-Thoughts-Q_8](https://huggingface.co/mlx-community/Qwen3-18B-A3B-Stranger-Thoughts-Q_8) was
converted to MLX format from [DavidAU/Qwen3-18B-A3B-Stranger-Thoughts](https://huggingface.co/DavidAU/Qwen3-18B-A3B-Stranger-Thoughts)
using mlx-lm version **0.25.2**.

## Use with mlx

```bash
pip install mlx-lm
```

```python
from mlx_lm import load, generate

model, tokenizer = load("mlx-community/Qwen3-18B-A3B-Stranger-Thoughts-Q_8")

prompt = "hello"

if tokenizer.chat_template is not None:
    messages = [{"role": "user", "content": prompt}]
    prompt = tokenizer.apply_chat_template(
        messages, add_generation_prompt=True
    )

response = generate(model, tokenizer, prompt=prompt, verbose=True)
```