AI News (2026/1/20): GLM-4.7-Flash Open-Sourced and Free
Executive Summary:
Zhipu AI has officially open-sourced and released the GLM-4.7-Flash model, which features a hybrid thinking architecture with a total of 30B parameters and only 3B active parameters, providing a high-performance option for lightweight deployment. The model performs excellently on mainstream benchmarks such as SWE-bench Verified and τ²-Bench, surpassing open-source models of the same size to achieve SOTA levels.
News Details
Zhipu AI announced on January 20th the official open-sourcing and release of the GLM-4.7-Flash model. This model employs a hybrid thinking architecture with a total of 30B parameters, but only 3B active parameters, significantly reducing computational resource requirements while maintaining high performance. GLM-4.7-Flash excels in multiple mainstream benchmarks, particularly in SWE-bench Verified and τ²-Bench, surpassing open-source models of the same size to achieve SOTA levels.
Key Points
Hybrid Thinking Architecture: GLM-4.7-Flash adopts a hybrid thinking architecture that dynamically adjusts the number of active parameters to optimize the use of computational resources. This architecture allows the model to maintain high performance while adapting to different hardware environments.
Lightweight Deployment: Despite having a total of 30B parameters, the active parameters are only 3B, enabling GLM-4.7-Flash to run efficiently on resource-constrained devices. This is particularly significant for applications in edge computing and mobile devices.
Benchmark Performance: GLM-4.7-Flash performs excellently in multiple mainstream benchmarks. In particular, it outperforms open-source models of the same size in SWE-bench Verified and τ²-Bench, achieving SOTA levels.
AI-ALL In-Depth Commentary
The open-sourcing of GLM-4.7-Flash marks a significant advancement in balancing lightweight and high-performance AI models. The introduction of the hybrid thinking architecture not only enhances the model's flexibility and adaptability but also significantly reduces deployment costs. This is particularly important for promoting the application of AI technology in edge computing and mobile devices. Additionally, GLM-4.7-Flash's excellent performance in multiple benchmarks further validates its technical advantages and is likely to become a new choice for developers and researchers.