OMIBench: Benchmarking Olympiad-Level Multi-Image Reasoning in Large Vision-Language Model
This provides a focused resource for studying multi-image reasoning in LVLMs, addressing a gap in existing benchmarks that emphasize single-image analysis.
The authors tackled the problem of evaluating Olympiad-level reasoning in large vision-language models when evidence is distributed across multiple images, and found that even the strongest models like Gemini-3-Pro achieve only about 50% accuracy on their new benchmark, OMIBench.
Large vision-language models (LVLMs) have made substantial advances in reasoning tasks at the Olympiad level. Nevertheless, current Olympiad-level multimodal reasoning benchmarks for these models often emphasize single-image analysis and fail to exploit contextual information across multiple images. We present OMIBench, a benchmark designed to evaluate Olympiad-level reasoning when the required evidence is distributed over multiple images. It contains problems from biology, chemistry, mathematics, and physics Olympiads, together with manually annotated rationales and evaluation protocols for both exact and semantic answer matching. Across extensive experiments on OMIBench, we observe meaningful performance gaps in existing models. Even the strongest LVLMs, such as Gemini-3-Pro, attain only about 50% on the benchmark. These results position OMIBench as a focused resources for studying and improving multi-image reasoning in LVLMs.