Abstract: In the intelligent evolution of lightweight structural design, structural topology optimization requires the integration of highly efficient auxiliary tools to streamline its solution and analysis methods, including code generation, load case adaptation, and post-processing. As powerful AI-assisted tools, mainstream Large Language Models (LLMs)—such as DeepSeek-V3, ChatGPT-5.2, and Gemini 3.0 Pro—can achieve automated generation of topology optimization code and computational reliability verification in the field of mechanics when combined with a standardized human-computer interaction prompting workflow. Benchmarked against the classic 99-line code for structural topology optimization design, this study investigates the interactive processes and correction methods for code generation across mainstream AI tools using typical numerical examples. Furthermore, it explores the capability migration effects and code generation boundaries of different AI tools, and presents extended case studies utilizing AI tools to assist in structural design. The research findings provide valuable insights and practical approaches for university students to leverage AI tools to facilitate their learning and research.