_version_ 1866916814268661760
author Gao, Yu
Guo, Haoyuan
Hoang, Tuyen
Huang, Weilin
Jiang, Lu
Kong, Fangyuan
Li, Huixia
Li, Jiashi
Li, Liang
Li, Xiaojie
Li, Xunsong
Li, Yifu
Lin, Shanchuan
Lin, Zhijie
Liu, Jiawei
Liu, Shu
Nie, Xiaonan
Qing, Zhiwu
Ren, Yuxi
Sun, Li
Tian, Zhi
Wang, Rui
Wang, Sen
Wei, Guoqiang
Wu, Guohong
Wu, Jie
Xia, Ruiqi
Xiao, Fei
Xiao, Xuefeng
Yan, Jiangqiao
Yang, Ceyuan
Yang, Jianchao
Yang, Runkai
Yang, Tao
Yang, Yihang
Ye, Zilyu
Zeng, Xuejiao
Zeng, Yan
Zhang, Heng
Zhao, Yang
Zheng, Xiaozheng
Zhu, Peihao
Zou, Jiaxin
Zuo, Feilong
author_facet Gao, Yu
Guo, Haoyuan
Hoang, Tuyen
Huang, Weilin
Jiang, Lu
Kong, Fangyuan
Li, Huixia
Li, Jiashi
Li, Liang
Li, Xiaojie
Li, Xunsong
Li, Yifu
Lin, Shanchuan
Lin, Zhijie
Liu, Jiawei
Liu, Shu
Nie, Xiaonan
Qing, Zhiwu
Ren, Yuxi
Sun, Li
Tian, Zhi
Wang, Rui
Wang, Sen
Wei, Guoqiang
Wu, Guohong
Wu, Jie
Xia, Ruiqi
Xiao, Fei
Xiao, Xuefeng
Yan, Jiangqiao
Yang, Ceyuan
Yang, Jianchao
Yang, Runkai
Yang, Tao
Yang, Yihang
Ye, Zilyu
Zeng, Xuejiao
Zeng, Yan
Zhang, Heng
Zhao, Yang
Zheng, Xiaozheng
Zhu, Peihao
Zou, Jiaxin
Zuo, Feilong
contents Notable breakthroughs in diffusion modeling have propelled rapid improvements in video generation, yet current foundational model still face critical challenges in simultaneously balancing prompt following, motion plausibility, and visual quality. In this report, we introduce Seedance 1.0, a high-performance and inference-efficient video foundation generation model that integrates several core technical improvements: (i) multi-source data curation augmented with precision and meaningful video captioning, enabling comprehensive learning across diverse scenarios; (ii) an efficient architecture design with proposed training paradigm, which allows for natively supporting multi-shot generation and jointly learning of both text-to-video and image-to-video tasks. (iii) carefully-optimized post-training approaches leveraging fine-grained supervised fine-tuning, and video-specific RLHF with multi-dimensional reward mechanisms for comprehensive performance improvements; (iv) excellent model acceleration achieving ~10x inference speedup through multi-stage distillation strategies and system-level optimizations. Seedance 1.0 can generate a 5-second video at 1080p resolution only with 41.4 seconds (NVIDIA-L20). Compared to state-of-the-art video generation models, Seedance 1.0 stands out with high-quality and fast video generation having superior spatiotemporal fluidity with structural stability, precise instruction adherence in complex multi-subject contexts, native multi-shot narrative coherence with consistent subject representation.
format Preprint
id arxiv_https___arxiv_org_abs_2506_09113
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Seedance 1.0: Exploring the Boundaries of Video Generation Models
Gao, Yu
Guo, Haoyuan
Hoang, Tuyen
Huang, Weilin
Jiang, Lu
Kong, Fangyuan
Li, Huixia
Li, Jiashi
Li, Liang
Li, Xiaojie
Li, Xunsong
Li, Yifu
Lin, Shanchuan
Lin, Zhijie
Liu, Jiawei
Liu, Shu
Nie, Xiaonan
Qing, Zhiwu
Ren, Yuxi
Sun, Li
Tian, Zhi
Wang, Rui
Wang, Sen
Wei, Guoqiang
Wu, Guohong
Wu, Jie
Xia, Ruiqi
Xiao, Fei
Xiao, Xuefeng
Yan, Jiangqiao
Yang, Ceyuan
Yang, Jianchao
Yang, Runkai
Yang, Tao
Yang, Yihang
Ye, Zilyu
Zeng, Xuejiao
Zeng, Yan
Zhang, Heng
Zhao, Yang
Zheng, Xiaozheng
Zhu, Peihao
Zou, Jiaxin
Zuo, Feilong
Computer Vision and Pattern Recognition
Notable breakthroughs in diffusion modeling have propelled rapid improvements in video generation, yet current foundational model still face critical challenges in simultaneously balancing prompt following, motion plausibility, and visual quality. In this report, we introduce Seedance 1.0, a high-performance and inference-efficient video foundation generation model that integrates several core technical improvements: (i) multi-source data curation augmented with precision and meaningful video captioning, enabling comprehensive learning across diverse scenarios; (ii) an efficient architecture design with proposed training paradigm, which allows for natively supporting multi-shot generation and jointly learning of both text-to-video and image-to-video tasks. (iii) carefully-optimized post-training approaches leveraging fine-grained supervised fine-tuning, and video-specific RLHF with multi-dimensional reward mechanisms for comprehensive performance improvements; (iv) excellent model acceleration achieving ~10x inference speedup through multi-stage distillation strategies and system-level optimizations. Seedance 1.0 can generate a 5-second video at 1080p resolution only with 41.4 seconds (NVIDIA-L20). Compared to state-of-the-art video generation models, Seedance 1.0 stands out with high-quality and fast video generation having superior spatiotemporal fluidity with structural stability, precise instruction adherence in complex multi-subject contexts, native multi-shot narrative coherence with consistent subject representation.
title Seedance 1.0: Exploring the Boundaries of Video Generation Models
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2506.09113