LLaVA-OneVision-1.5: Fully Open Framework for Democratized Multimodal
  Training

Xiang An; Yin Xie; Kaicheng Yang; Wenkang Zhang; Xiuwei Zhao; Zheng Cheng; Yirui Wang; Songcen Xu; Changrui Chen; Chunsheng Wu; Huajie Tan; Chunyuan Li; Jing Yang; Jie Yu; Xiyao Wang; Bin Qin; Yumeng Wang; Zizhen Yan; Ziyong Feng; Ziwei Liu; Bo Li; Jiankang Deng

Paper · arXiv 2509.23661

LLaVA-OneVision-1.5: Fully Open Framework for Democratized Multimodal Training

Xiang An, Yin Xie, Kaicheng Yang, Wenkang Zhang, Xiuwei Zhao, Zheng Cheng, Yirui Wang, Songcen Xu, Changrui Chen, Chunsheng Wu, Huajie Tan, Chunyuan Li, Jing Yang, Jie Yu, Xiyao Wang, Bin Qin, Yumeng Wang, Zizhen Yan, Ziyong Feng, Ziwei Liu, Bo Li, Jiankang Deng

52 upvotesSeptember 28, 2025arXiv 预印本

AI 摘要

LLaVA-OneVision-1.5 is a family of large multimodal models that achieves state-of-the-art performance with reduced costs through efficient training and high-quality datasets.

Large Multimodal ModelsLMMsLarge-Scale Curated Datasetspretraining datasetinstruction datasetmultimodal tokensefficient training frameworkoffline parallel data packing strategystate-of-the-art performancedownstream tasksQwen2.5-VL-7BQwen2.5-VL-3B

Abstract

We present LLaVA-OneVision-1.5, a novel family of Large Multimodal Models (LMMs) that achieve state-of-the-art performance with significantly reduced computational and financial costs. Different from the existing works, LLaVA-OneVision-1.5 provides an open, efficient, and reproducible framework for building high-quality vision-language models entirely from scratch. The LLaVA-OneVision-1.5 release comprises three primary components: (1) Large-Scale Curated Datasets: We construct an 85M concept-balanced pretraining dataset LLaVA-OneVision-1.5-Mid-Traning and a meticulously curated 26M instruction dataset LLaVA-OneVision-1.5-Instruct, collectively encompassing 64B compressed multimodal tokens. (2) Efficient Training Framework: We develop a complete end-to-end efficient training framework leveraging an offline parallel data packing strategy to facilitate the training of LLaVA-OneVision-1.5 within a $16,000 budget. (3) State-of-the-art Performance: Experimental results demonstrate that LLaVA-OneVision1.5 yields exceptionally competitive performance across a broad range of downstream tasks. Specifically, LLaVA-OneVision-1.5-8B outperforms Qwen2.5-VL-7B on 18 of 27 benchmarks, and LLaVA-OneVision-1.5-4B surpasses Qwen2.5-VL-3B on all 27 benchmarks. We anticipate releasing LLaVA-OneVision-1.5-RL shortly and encourage the community to await further updates.