MiniCPM-V 4.5: Cooking Efficient MLLMs via Architecture, Data, and Training Recipe

Tianyu Yu, Zefan Wang, chongyi, Fuwei Huang, Wenshuo Ma, Zhihui He, Tianchi Cai, weize, Yuxiang Huang, Yuanqian Zhao, Bokai Xu, Cuiunbo, Yingjing Xu, Liqing Ruan, zly, Hanyu Liu, Jingkun Tang, Hongyuan Liu, Qining Guo, wenhaohu, Bingxiang He, Jie Zhou, Jie Cai, Ji Qi, Zonghao Guo, Chi Chen, Guoyang Zeng, Yuxuan Li, Ganqu Cui, Ning Ding, Xu Han, Yuan Yao, Zhiyuan Liu, Maosong Sun

MiniCPM-V 4.5: Cooking Efficient MLLMs via Architecture, Data, and Training Recipe: 63 upvotes on Hugging Face Daily Papers, #4 of 24 papers on 2025-09-24. Day-by-day upvote history.

Multimodal Large Language Models (MLLMs) are undergoing rapid progress and represent the frontier of AI development. However, their training and inference efficiency have emerged as a core bottleneck in making MLLMs more accessible and scalable. To address the challenges, we present MiniCPM-V 4.5, an 8B parameter model designed for high efficiency and strong performance. We introduce three core improvements in model architecture, data strategy and training method: a unified 3D-Resampler model architecture for highly compact encoding over images and videos, a unified learning paradigm for document knowledge and text recognition without heavy data engineering, and a hybrid reinforcement learning strategy for proficiency in both short and long reasoning modes. Comprehensive experimental results in OpenCompass evaluation show that MiniCPM-V 4.5 surpasses widely used proprietary models such as GPT-4o-latest, and significantly larger open-source models such as Qwen2.5-VL 72B. Notably, the strong performance is achieved with remarkable efficiency. For example, on the widely adopted VideoMME benchmark, MiniCPM-V 4.5 achieves state-of-the-art performance among models under 30B size, using just 46.7\% GPU memory cost and 8.7\% inference time of Qwen2.5-VL 7B.

Paper page on Hugging Face · arXiv

Data: hysts-bot-data/daily-papers-stats and the Daily Papers API. Open data: tardellirs/paper-pulse-data. Sister project: Model Pulse, the download history of every model on the Hub.