GLM-4.7

GLM-4.7 is a foundation model engineered to serve as a versatile partner for intelligent agents and software development. With a architecture comprising 355 billion total parameters and 32 billion active parameters, the model is designed to unify reasoning, coding, and tool-use capabilities. It excels in multi-step production environments, offering improved consistency and stability when managing lengthy task cycles. Beyond its technical core, the model features a specialized focus on UI quality, enabling it to generate cleaner, more modern webpages and well-structured presentation slides.

Built upon the lineage of the GLM-4.6, this iteration introduces significant advancements in multilingual agentic coding and terminal-based operations, as evidenced by performance gains on benchmarks like SWE-bench and Terminal Bench 2.0. The model supports sophisticated thinking-before-acting processes, allowing it to integrate seamlessly into mainstream agent frameworks. Its refined reasoning capabilities extend to mathematical problem-solving and complex logic, while its enhanced tool-using proficiency ensures reliable performance across web browsing and external application interactions. These improvements make it a robust choice for developers seeking a consistent, high-performing system for real-world production workflows.