Journal article · 2026 · Open access

Multimodal learning with next-token prediction for large multimodal models

Xinlong Wang, Yufeng Cui, Jinsheng Wang, Fan Zhang, Yueze Wang, Xiaosong Zhang, Zhengxiong Luo, Quan Sun, Zhen Li, Yuqi Wang, Qiying Yu, Yingli Zhao and 14 more

Nature, 650(8101), 327–333 · Published 28 January 2026 · DOI 10.1038/s41586-025-10041-x

Open full textCite this paper

In brief

“Multimodal learning with next-token prediction for large multimodal models” is a 2026 journal article by Xinlong Wang and 25 co-authors, published in Nature. OpenAlex counts 37 citations as of 16 November 2026, which puts it in the top 1% of works published in 2026 by citations. It is open access under a CC BY licence: the full text is free to read at Nature.

Citations
37
Most cited in
2026 (37)
Field-weighted impact
54.33
References
33
Authors
26

Abstract

Developing a unified algorithm that can learn from and generate across modalities such as text, images and video has been a fundamental challenge in artificial intelligence. Although next-token prediction has driven major advances in large language models1, its extension to multimodal domains has remained limited, and diffusion models for image and video synthesis2,3 and compositional frameworks that integrate vision encoders with language models4 still dominate. Here we introduce Emu3, a family of multimodal models trained solely with next-token prediction. Emu3 equals the performance of well-established task-specific models across both perception and generation, matching flagship systems while removing the need for diffusion or compositional architectures. It further demonstrates coherent, high-fidelity video generation, interleaved vision–language generation and vision–language–action modelling for robotic manipulation. By reducing multimodal learning to unified token prediction, Emu3 establishes a robust foundation for large-scale multimodal modelling and offers a promising route towards unified multimodal intelligence. Emu3 enables large-scale text, image and video learning based solely on next-token prediction, matching the generation and perception performance of task-specific methods, with implications for the development of scalable and unified multimodal intelligence systems.

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

Wang, X., Cui, Y., Wang, J., Zhang, F., Wang, Y., Zhang, X., Luo, Z., Sun, Q., Li, Z., Wang, Y., Yu, Q., Zhao, Y., Ao, Y., Min, X., Men, C., Wu, B., Zhao, B., Zhang, B., Wang, L., . . . Huang, T. (2026). Multimodal learning with next-token prediction for large multimodal models. Nature, 650(8101), 327–333. https://doi.org/10.1038/s41586-025-10041-x

MLA 9

Wang, Xinlong, et al. “Multimodal learning with next-token prediction for large multimodal models.” Nature, vol. 650, no. 8101, 2026, pp. 327–333. https://doi.org/10.1038/s41586-025-10041-x.

Chicago

Wang, Xinlong, Yufeng Cui, Jinsheng Wang, Fan Zhang, Yueze Wang, Xiaosong Zhang, Zhengxiong Luo, et al. 2026. “Multimodal learning with next-token prediction for large multimodal models.” Nature 650 (8101): 327–333. https://doi.org/10.1038/s41586-025-10041-x.

Harvard

Wang, X. et al. (2026) ‘Multimodal learning with next-token prediction for large multimodal models’, Nature, 650(8101), pp. 327–333. doi:10.1038/s41586-025-10041-x.

IEEE

X. Wang et al., “Multimodal learning with next-token prediction for large multimodal models,” Nature, vol. 650, no. 8101, pp. 327–333, 2026, doi: 10.1038/s41586-025-10041-x.

BibTeX

@article{wang2026multimodal,
  title     = {{Multimodal learning with next-token prediction for large multimodal models}},
  author    = {Wang, Xinlong and Cui, Yufeng and Wang, Jinsheng and Zhang, Fan and Wang, Yueze and Zhang, Xiaosong and Luo, Zhengxiong and Sun, Quan and Li, Zhen and Wang, Yuqi and Yu, Qiying and Zhao, Yingli and Ao, Yulong and Min, Xuebin and Men, Chunlei and Wu, Boya and Zhao, Bo and Zhang, Bowen and Wang, Liangdong and Liu, Guang and He, Zheqi and Yang, Xi and Liu, Jianxing and Lin, Yonghua and Wang, Zhongyuan and Huang, Tiejun},
  journal   = {Nature},
  year      = {2026},
  volume    = {650},
  number    = {8101},
  pages     = {327--333},
  doi       = {10.1038/s41586-025-10041-x},
  url       = {https://doi.org/10.1038/s41586-025-10041-x}
}

Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)

Topic and keywords

Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Multimodal Machine Learning Applications

  • multimodal learning
  • next-token prediction
  • large multimodal models
  • video generation
  • robotic manipulation

Work with this paper

Most-cited papers in Computer Science

All highly cited Computer Science papers

Questions about this paper

How do I cite this paper in APA style?

Wang, X., Cui, Y., Wang, J., Zhang, F., Wang, Y., Zhang, X., Luo, Z., Sun, Q., Li, Z., Wang, Y., Yu, Q., Zhao, Y., Ao, Y., Min, X., Men, C., Wu, B., Zhao, B., Zhang, B., Wang, L., . . . Huang, T. (2026). Multimodal learning with next-token prediction for large multimodal models. Nature, 650(8101), 327–333. https://doi.org/10.1038/s41586-025-10041-x

How many times has this paper been cited?

37 times, according to OpenAlex on 16 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.

Is this paper free to read?

Yes. It is open access at Nature: sign in and open the full text from this page.

Who wrote this paper?

Xinlong Wang, Yufeng Cui, Jinsheng Wang, Fan Zhang, Yueze Wang, Xiaosong Zhang, Zhengxiong Luo, Quan Sun, Zhen Li, Yuqi Wang, Qiying Yu, Yingli Zhao and 14 more.

What is the DOI of this paper?

10.1038/s41586-025-10041-x, permanently at https://doi.org/10.1038/s41586-025-10041-x. Sign in and open it from this page.

Bibliographic data from OpenAlex (CC0), last checked 16 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.