T2I-ReasonBench: Benchmarking Reasoning-Informed Text-to-Image Generation

This paper introduces T2I-ReasonBench, a novel benchmark designed to explore the reasoning border of T2I models. T2I-ReasonBench comprises 800 meticulously designed prompts organized into four dimensions: (1) Idiom Interpretation, (2) Textual Image Design, (3) Entity-Reasoning, and (4) Scientific-Reasoning. These dimensions challenge models to infer latent meaning, integrate domain knowledge, and resolve contextual ambiguities. To quantify the performance, we introduce a two-stage evaluation framework: a large language model (LLM) generates prompt-specific question-criterion pairs that evaluate if the image includes the essential elements resulting from correct reasoning; a multimodal LLM (MLLM) then scores the generated image against these criteria. Experiments across 14 state-of-the-art T2I models reveal that open-source models exhibit critical limitations in reasoning-informed generation, while proprietary models like GPT-Image-1 demonstrate stronger reasoning and knowledge integration. Our findings underscore the necessity to improve reasoning capabilities in next-generation T2I systems. This work provides a foundational benchmark and evaluation protocol to guide future research towards reasoning-informed T2I synthesis.