Qwen3.6-35B-A3B
Qwen3.6-35B-A3B is an efficient mixture-of-experts (MoE) model with 35B total parameters and only 3B active. It delivers strong agentic coding performance, significantly outperforming Qwen3.5-35B-A3B and competing with larger dense models like Qwen3.5-27B and Gemma4-31B. Supporting both multimodal thinking and non-thinking modes, it is a highly versatile open-source model, now available on Qwen Studio, via API, and as open weights.
Input
Output
Qwen3.6-35B-A3B Documentation
Qwen3.6-35B-A3B is a sparse Mixture-of-Experts (MoE) foundation model. While it possesses a total capacity of 35.1 Billion parameters, its dynamic routing mechanism activates only 3.2 Billion parameters for any given input. This design provides a "Best of Both Worlds" solution: the vast knowledge of a large model with the rapid response of a small one.
Key Capabilities
- Intelligence Density: Surpasses previous 7B and 14B dense models in logical reasoning, math, and coding benchmarks.
- 256K Context Window: Native support for processing entire code repositories or long legal documents.
- Multilingual Mastery: Fine-tuned on over 30 trillion tokens, supporting native-level fluency in 29+ languages.
- Agentic Efficiency: Optimized for tool-calling and long-range instruction following, making it ideal for autonomous AI agents.
- Low VRAM Footprint: Despite being a "35B" model, its sparse nature and optimized kernels make it accessible to mid-range hardware.
Request Parameters
To interact with this model via the us-01.bytecompute.ai endpoint:
| Parameter | Type | Required | Description |
|---|---|---|---|
model |
string |
Yes | Use "Qwen3.6-35B-A3B". |
messages |
array |
Yes | Standard chat array (system, user, assistant). |
max_tokens |
integer |
No | Maximum generation length. Default: 4096. |
temperature |
float |
No | 0.1 for coding/logic; 0.8 for creative tasks. |
top_p |
float |
No | Nucleus sampling threshold. Default: 0.9. |
Unlock the most affordable AI hosting
Run models at scale with our fully managed GPU infrastructure, delivering enterprise-grade uptime at the industry's best rates.