---
title: MLX-Qwen3.5-9B-DeepSeek-V4-Flash-bf16
canonical_url: "https://www.modelscope.cn/models/Jackrong/MLX-Qwen3.5-9B-DeepSeek-V4-Flash-bf16"
md_url: "https://www.modelscope.cn/models/Jackrong/MLX-Qwen3.5-9B-DeepSeek-V4-Flash-bf16.md"
repository: Jackrong/MLX-Qwen3.5-9B-DeepSeek-V4-Flash-bf16
last_updated: 2026-04-30
license: apache-2.0
pipeline_tag: text-generation
tasks:
  - text-generation
model_type:
  - qwen3_5
architectures:
  - Qwen3_5ForConditionalGeneration
base_model:
  - Jackrong/Qwen3.5-9B-DeepSeek-V4-Flash
base_model_relation: finetune
parameters: 9.0B
tensor_type:
  - BF16
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
language:
  - en
  - zh
  - ko
  - ja
  - es
  - ru
downloads: 168
stars: 1
tags:
  - text-generation-inference
  - transformers
  - unsloth
  - qwen3_5
  - reasoning
  - distillation
  - deepseek
  - deepseek-v4
  - sft
  - long-cot
  - chain-of-thought
  - efficient-inference
  - agent
  - multilingual
  - mlx
---

# MLX-Qwen3.5-9B-DeepSeek-V4-Flash-bf16

> MLX-Qwen3.5-9B-DeepSeek-V4-Flash-bf16 - Jackrong 在 ModelScope 开源的模型。Jackrong/MLX-Qwen3.5-9B-DeepSeek-V4-Flash-bf16

Jackrong/MLX-Qwen3.5-9B-DeepSeek-V4-Flash-bf16 是 ModelScope 魔搭社区上的 9.0B 参数text-generation模型，采用 apache-2.0 许可，基于 Jackrong/Qwen3.5-9B-DeepSeek-V4-Flash 构建。

- **Repository**: Jackrong/MLX-Qwen3.5-9B-DeepSeek-V4-Flash-bf16
- **License**: apache-2.0
- **Tasks**: text-generation
- **Parameters**: 9.0B
- **Base model**: Jackrong/Qwen3.5-9B-DeepSeek-V4-Flash
- **Tags**: text-generation-inference, transformers, unsloth, qwen3_5, reasoning, distillation, deepseek, deepseek-v4, sft, long-cot, chain-of-thought, efficient-inference, agent, multilingual, mlx
- **Downloads**: 168
- **Stars**: 1
- **Last updated**: 2026-04-30

Source: https://www.modelscope.cn/models/Jackrong/MLX-Qwen3.5-9B-DeepSeek-V4-Flash-bf16

---

# Jackrong/MLX-Qwen3.5-9B-DeepSeek-V4-Flash-bf16

This model [Jackrong/MLX-Qwen3.5-9B-DeepSeek-V4-Flash-bf16](https://huggingface.co/Jackrong/MLX-Qwen3.5-9B-DeepSeek-V4-Flash-bf16) was
converted to MLX format from [Jackrong/Qwen3.5-9B-DeepSeek-V4-Flash](https://huggingface.co/Jackrong/Qwen3.5-9B-DeepSeek-V4-Flash)
using mlx-lm version **0.30.7**.

## Use with mlx

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

```python
from mlx_lm import load, generate

model, tokenizer = load("Jackrong/MLX-Qwen3.5-9B-DeepSeek-V4-Flash-bf16")

prompt = "hello"

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

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