AI News (2026/2/4): Qwen3-Coder-Next Open-Sourced: Small and Powerful!
Executive Summary:
The Tongyi large model has announced the open-sourcing of the Qwen3-Coder-Next programming AI model. This model uses an 80B parameter MoE (Mixture of Experts) architecture, with only 3B parameters activated during each inference, significantly reducing computational requirements. On the SWE-Bench Verified benchmark, Qwen3-Coder-Next achieved a problem-solving rate of over 70%, performing on par with larger dense models.
Tongyi Qwen3-Coder-Next Open-Sourced: Small and Powerful!
News Details
Tongyi large model announced on February 4th the open-sourcing of the Qwen3-Coder-Next programming AI model. This model is designed to provide developers with an efficient, low-computational-demand programming assistant. Qwen3-Coder-Next uses an 80B total parameter MoE (Mixture of Experts) architecture, with only 3B parameters activated during each inference, significantly reducing the consumption of computing resources. On the SWE-Bench Verified benchmark, Qwen3-Coder-Next demonstrated outstanding performance, achieving a problem-solving rate of over 70%, comparable to larger dense models.
Key Points
MoE Architecture: Qwen3-Coder-Next employs an 80B total parameter MoE architecture, which drastically reduces the demand for computing resources by activating only 3B parameters during each inference. This architecture not only enhances the model's efficiency but also maintains a high performance level.
Problem-Solving Rate: On the SWE-Bench Verified benchmark, Qwen3-Coder-Next achieved a problem-solving rate of over 70%. This result indicates that despite its smaller size, the model performs exceptionally well in practical programming tasks.
Open-Source Scope: Tongyi large model has fully open-sourced Qwen3-Coder-Next, allowing developers to freely access and use the model. Additionally, Tongyi provides detailed documentation and sample code to help developers quickly get started and integrate the model into their existing development environments.
AI-ALL In-Depth Commentary
The open-sourcing of Qwen3-Coder-Next marks a new balance between efficiency and performance in AI programming assistant tools. The application of MoE architecture not only reduces computational requirements but also offers performance comparable to large-scale dense models. This is particularly important for development teams with limited resources, as they can perform efficient programming work with fewer computing resources without sacrificing performance. Furthermore, the fully open-source strategy helps drive community innovation and technological advancement, accelerating the application and development of AI in the programming field.