MTabVQA: Evaluating Multi-Tabular Reasoning of Language Models in Visual Space

Vision-Language Models (VLMs) have demonstrated remarkable capabilities in interpreting visual layouts and text. However, a significant challenge remains in their ability to interpret and reason robustly over multi-tabular data presented as images, a common occurrence in real-world scenarios such as web pages and digital documents. Existing benchmarks typically address single tables or non-visual data (text/structured). This leaves a critical gap: they do not assess the ability to parse diverse table images, correlate information across them, and perform multi-hop reasoning on the combined visual data. We introduce MTab-VQA, a novel benchmark specifically designed for multi-tabular visual question answering to bridge this gap. MTabVQA comprises 3,745 complex question-answer pairs that necessitate multi-hop reasoning across several visually rendered table images. We provide extensive benchmark results for state-of-the-art VLMs on MTabVQA, revealing significant performance limitations. We further investigate posttraining techniques to enhance the multitabular reasoning abilities of vision-language models and release MTabVQA-Instruct, a large-scale instruction-tuning dataset. Our experiments show that fine-tuning VLMs with MTabVQA-Instruct substantially improves their performance on visual multi-tabular reasoning. Code and dataset are available online 1 .