Effected time

2025-06-30 12:00:00 (UTC+08) [Subject to the actual update time]

Effective time

Alibaba Cloud Model Studio will update its Qwen-Plus and Qwen-Turbo models on June 30, 2025.

Impact

After the update:

Qwen-Plus: Updated to the Qwen3 series Plus model. It effectively integrates thinking mode and non-thinking modes, allowing mode switching during conversations. Its reasoning capability significantly surpasses QwQ, and its general capability significantly exceeds Qwen2.5-Plus, reaching the SOTA level among models of the similar scale within the industry.

Qwen-Turbo: Updated to the Qwen3 series Plus model. It effectively integrates thinking mode and non-thinking modes, allowing mode switching during conversations. Its reasoning capability matches QwQ-32B with a smaller parameter scale, and its general capability significantly surpasses Qwen2.5-Turbo, reaching the SOTA level among models of the similar scale within the industry.

The Qwen3 series models support both thinking and non-thinking modes. You can use the enable_thinking parameter to switch between them. In thinking mode, only streaming output is supported. {Code samples}

We have already pre-released the updated versions: Qwen-Plus-2025-04-28 and Qwen-Turbo-2025-04-28, which you can preview now. If you wish to stay at the current model, manually switch to Qwen-Plus-2025-01-25 or Qwen-Turbo-2024-11-01 before June 30. (Note: These snapshots are not intended for long-term use.)

Model mapping:

qwen-plus: Before the update, its performance is equivalent to qwen-plus-2025-01-25. After the update, it will be equivalent to qwen-plus-2025-04-28.

qwen-turbo: Before the update, its performance is equivalent to qwen-turbo-2024-11-01. After the update, it will be equivalent to qwen-turbo-2025-04-28.

Go to the {Model Studio console}