FullFront: Benchmarking MLLMs Across the Full Front-End Engineering Workflow
FullFront introduces a benchmark to test Multimodal Large Language Models (MLLMs) across the entire front-end engineering process. It evaluates models on design, perception, and code generation tasks using both visual and code metrics. The results provide a robust assessment of an MLLM's capability in translating visual concepts into functional web elements.