<![CDATA[In the context of modern architectural design that demands innovation, speed, and efficiency, the emergence of generative artificial intelligence (AI) introduces a new paradigm in the creative process. This technology enables architects to explore design ideas more rapidly and extensively through diffusion-based algorithms capable of producing complex architectural visuals in a short amount of time. This study aims to empirically evaluate the effectiveness and efficiency of generative AI models, particularly Stable Diffusion v2.1, in supporting the stages of ideation, sketching, and architectural modeling. The research employs both qualitative and quantitative approaches through a comparative experiment between manual design and AI-assisted design. Measurements were conducted using four main parameters: production time, visual complexity, rendering sharpness, and the number of design iterations. The results indicate that the generative AI model can accelerate production time by up to 35% greater efficiency compared to the manual method. Furthermore, the Visual Complexity Score (VCS) reached 8.5/10 for AI-generated designs and 6.2/10 for manual ones, with an increase in rendering resolution up to 450 PPI. However, limitations were observed in semantic interpretation and the model’s dependence on well-crafted prompts. This study concludes that the integration of generative AI in architectural design not only enhances the efficiency and effectiveness of the design process but also expands the creative potential of architects. The research contributes to the development of sustainable digital architecture and supports the achievement of SDG 9 (Industry, Innovation, and Infrastructure) and SDG 11 (Sustainable Cities and Communities). ]]>
