The Qwen3.5 122B-A10B native vision-language model is built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency. In terms of overall performance, this model is second only to Qwen3.5-397B-A17B. Its text capabilities significantly outperform those of Qwen3-235B-2507, and its visual capabilities surpass those of Qwen3-VL-235B.
Inference Service Provider
The inference service provider for qwen3.5-122b-a10b is Alibaba Cloud Model Studio.
Model Capabilities
- China (Beijing)
- Singapore
- Germany (Frankfurt)
- US (Virginia)
| Capability | Support | Capability | Support |
|---|---|---|---|
Input Modality | Text Image Video | Output Modality | Text |
Model Experience | Supported | Function Calling | Supported |
Structured Outputs | Supported | Web Search | Supported |
Prefix Completion | Supported | Context Caching | Unsupported |
Batch Inference | Unsupported | Fine-tuning | Unsupported |
Context Limits
| Parameter | Value | Parameter | Value |
|---|---|---|---|
Max Input Length | 260096 | Max Output Length | 65536 |
Context Window | 262144 | Max Input Length (Thinking Mode) | 258048 |
Max Output Length (Thinking Mode) | 65536 | Max Chain-of-Thought Length | 81920 |
Pricing
This page only shows the original pricing for model API calls, excluding any limited-time promotions. Visit Model Studio Console for promotional offers.
- China (Beijing)
- Singapore
- Germany (Frankfurt)
- US (Virginia)
Input<=128k
Input<=128k
| Billing Item | Price (USD) | Unit |
|---|---|---|
Input | 0.115 | Per 1M tokens |
Output | 0.917 | Per 1M tokens |
128k<Input<=256k
128k<Input<=256k
| Billing Item | Price (USD) | Unit |
|---|---|---|
Input | 0.287 | Per 1M tokens |
Output | 2.294 | Per 1M tokens |
Rate Limits
- China (Beijing)
- Singapore
- Germany (Frankfurt)
- US (Virginia)
| Parameter | Value |
|---|---|
RPM (Requests Per Minute) | 600 |
TPM (Tokens Per Minute) | 1,000,000 |