ShareGPT-4o-Image is a large-scale, high-quality image generation dataset.ShareGPT-4o-Image: Aligning Multimodal Models with GPT-4o-Level Image Generation", which aims to migrate GPT-4o-level image generation capabilities to open source multimodal models. All images in this dataset are generated by GPT-4o's image generation function. The data contains a total of 92,256 image generation samples from GPT-4o, including 45,717 text-to-image and 46,539 text-and-image-to-image prompts, all carefully selected to ensure diversity and quality. This dataset covers a wide range of styles and embodied visual reasoning scenarios, while reflecting GPT-4o's advantages in instruction following and visual aesthetics.

Citation

@misc{chen2025sharegpt4oimg,
title={ShareGPT-4o-Image: Aligning Multimodal Models with GPT-4o-Level Image Generation},
author={Junying Chen and Zhenyang Cai and Pengcheng Chen and Shunian Chen and Ke Ji and Xidong Wang and Yunjin Yang and Benyou Wang},
year={2025},
eprint={2506.18095},
archivePrefix={arXiv},
primaryClass={cs.CV},
url={https://arxiv.org/abs/2506.18095},
}