# Qwen3.8-Max

Qwen3.8-Max is a 2.4-trillion-parameter Mixture-of-Experts (MoE) flagship model. It sets a new bar for autonomous coding, professional co-working, long-horizon planning, and native multimodal intelligence.

<p align="center">
  <a href="https://qwen.ai/blog?id=qwen3.8">📖 Blog</a> •
  <a href="https://chat.qwen.ai/?models=qwen3.8-max">✅ Qwen Studio</a> •
  <a href="https://qwencloud.com/models/qwen3.8-max">⚡ API Reference</a> •
  <a href="https://click.alibabacloud.com/m/20000000960/">☁️ Alibaba Cloud Model Studio</a> • 
  <a href="https://click.alibabacloud.com/m/20000000960/">⚡ Qwen Cloud</a>
</p>

---

## 🚀 Key Features

* **Autonomous Coding**: 10+ days of self-evolving development from an empty folder to production without hand-holding. View the complete project trace at [qwen-code-dev-bot/oh-my-cli](https://github.com/qwen-code-dev-bot/oh-my-cli).
* **Production-Quality Deliverables**: Handles hundreds of specialized tasks across legal, financial, design, and other professional domains.
* **Long-Horizon Mastery**: System-level autonomous planning with closed-loop adaptive learning, driving 500+ turns of chip design optimization.
* **Native Multimodal Intelligence**: Vision functions as a continuous feedback loop for planning, execution, and self-correction.

---

## 📊 Evaluation & Performance
<img width="900" height="662" alt="HOw1F2ebgAAkpJJ" src="https://github.com/user-attachments/assets/2d52339a-9c6b-43b0-aa6e-f514050f6ba6" />



###  LM Performance Chart
<img width="1105" height="1689" alt="LM Performance" src="https://github.com/user-attachments/assets/52f25655-7d89-4108-aeee-7c529dd551ae" />

###  VL Performance Chart
<img width="750" height="1677" alt="VL Performance" src="https://github.com/user-attachments/assets/14bba8ef-9f13-4159-8bb1-67832f772dd5" />

---

## 💰 Pricing & Context Limits

### API Pricing

| Token Type | Price per 1M Tokens |
| :--- | :--- |
| **Input** | \$2.00 |
| **Output** | \$6.00 |
| **Implicit Caching** | \$0.25 |

### Context Limits
* **Max Context Window**: 1M tokens.
* **Max Input / Output**: 991.80K / 131.07K tokens.