ChartQAPro: A More Diverse and Challenging Benchmark for Chart Question Answering
Ahmed Masry, Mohammed Saidul Islam, Mahir Ahmed, Aayush Bajaj, Firoz Kabir, Aaryaman Kartha, Md Tahmid Rahman Laskar, Mizanur Rahman, Shadikur Rahman, Mehrad Shahmohammadi, Megh Thakkar, Md Rizwan Parvez, Enamul Hoque, Shafiq Joty
TL;DR
ChartQAPro substantially extends chart QA benchmarking by compiling 1341 real-world charts from 157 sources and 1948 human-verified QA pairs across diverse question types and multi-chart layouts. The authors evaluate 21 models, including closed- and open-source LVLMs and chart-specific systems, under direct, CoT, and PoT prompting, revealing large performance gaps relative to human baselines. Findings show LVLMs struggle with visual perception, multi-step reasoning, and context handling, with performance drops even for state-of-the-art models on ChartQAPro compared to prior benchmarks. Through detailed error analyses and ablations across chart types, answer types, and context, the work identifies concrete directions for improving multimodal chart understanding. The dataset and evaluation prompts are released to foster ongoing progress in robust real-world chart comprehension.
Abstract
Charts are ubiquitous, as people often use them to analyze data, answer questions, and discover critical insights. However, performing complex analytical tasks with charts requires significant perceptual and cognitive effort. Chart Question Answering (CQA) systems automate this process by enabling models to interpret and reason with visual representations of data. However, existing benchmarks like ChartQA lack real-world diversity and have recently shown performance saturation with modern large vision-language models (LVLMs). To address these limitations, we introduce ChartQAPro, a new benchmark that includes 1,341 charts from 157 diverse sources, spanning various chart types, including infographics and dashboards, and featuring 1,948 questions in various types, such as multiple-choice, conversational, hypothetical, and unanswerable questions, to better reflect real-world challenges. Our evaluations with 21 models show a substantial performance drop for LVLMs on ChartQAPro; e.g., Claude Sonnet 3.5 scores 90.5% on ChartQA but only 55.81% on ChartQAPro, underscoring the complexity of chart reasoning. We complement our findings with detailed error analyses and ablation studies, identifying key challenges and opportunities for advancing LVLMs in chart understanding and reasoning. We release ChartQAPro at https://github.com/vis-nlp/ChartQAPro.