VQ-VA World: Towards High-Quality Visual Question-Visual Answering
Abstract
This paper studies Visual Question-Visual Answering (VQ-VA): generating an image, rather than text, in response to a visual question -- an ability that has recently emerged in proprietary systems such as NanoBanana and GPT-Image. To also bring this capability to open-source models, we introduce VQ-VA World, a data-centric framework built around an agentic pipeline for large-scale, targeted data construction. Leveraging web-scale deployment, this pipeline crawls a massive amount of ~1.8M high-quality, interleaved image-text samples for model training. For evaluation, we further release IntelligentBench, a human-curated benchmark that systematically assesses VQ-VA along the aspects of world knowledge, design knowledge, and reasoning. Training with VQ-VA World data yields strong empirical gains: it helps LightFusion attain 53.06 on IntelligentBench, substantially surpassing the best prior open-source baselines (i.e., 7.78 from vanilla LightFusion; 1.94 from UniWorld-V1), and significantly narrowing the gap toward leading proprietary systems (e.g., 81.67 from NanoBanana; 82.64 from GPT-Image). By releasing the full suite of model weights, datasets, and pipelines, we hope to stimulate future research on VQ-VA.
Cite
@article{arxiv.2511.20573,
title = {VQ-VA World: Towards High-Quality Visual Question-Visual Answering},
author = {Chenhui Gou and Zilong Chen and Zeyu Wang and Feng Li and Deyao Zhu and Zicheng Duan and Kunchang Li and Chaorui Deng and Hongyi Yuan and Haoqi Fan and Cihang Xie and Jianfei Cai and Hamid Rezatofighi},
journal= {arXiv preprint arXiv:2511.20573},
year = {2025}
}