Journal article · 2025 · Open access

DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning

Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Peiyi Wang, Qihao Zhu, Runxin Xu, Ruoyu Zhang, Shirong Ma, Xiao Bi, Xiaokang Zhang, Xingkai Yu and 182 more

Nature, 645(8081), 633–638 · Published 17 September 2025 · DOI 10.1038/s41586-025-09422-z

Free PDFView at the publisherCite this paper

In brief

“DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning” is a 2025 journal article by Daya Guo and 193 co-authors, published in Nature. OpenAlex counts 16,903 citations as of 15 November 2026. It is open access under a CC BY licence: the full text is free to read at Nature.

Citations
16,903
Most cited in
2026 (9,565)
Field-weighted impact
4113.79
References
43
Authors
194

Abstract

General reasoning represents a long-standing and formidable challenge in artificial intelligence (AI). Recent breakthroughs, exemplified by large language models (LLMs) 1,2 and chain-of-thought (CoT) prompting 3 , have achieved considerable success on foundational reasoning tasks. However, this success is heavily contingent on extensive human-annotated demonstrations and the capabilities of models are still insufficient for more complex problems. Here we show that the reasoning abilities of LLMs can be incentivized through pure reinforcement learning (RL), obviating the need for human-labelled reasoning trajectories. The proposed RL framework facilitates the emergent development of advanced reasoning patterns, such as self-reflection, verification and dynamic strategy adaptation. Consequently, the trained model achieves superior performance on verifiable tasks such as mathematics, coding competitions and STEM fields, surpassing its counterparts trained through conventional supervised learning on human demonstrations. Moreover, the emergent reasoning patterns exhibited by these large-scale models can be systematically used to guide and enhance the reasoning capabilities of smaller models.

Cite this paper

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

APA 7

Guo, D., Yang, D., Zhang, H., Song, J., Wang, P., Zhu, Q., Xu, R., Zhang, R., Ma, S., Bi, X., Zhang, X., Yu, X., Wu, Y., Wu, Z. F., Gou, Z., Shao, Z., Li, Z., Gao, Z., Liu, A., . . . Zhang, Z. (2025). DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning. Nature, 645(8081), 633–638. https://doi.org/10.1038/s41586-025-09422-z

MLA 9

Guo, Daya, et al. “DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning.” Nature, vol. 645, no. 8081, 2025, pp. 633–638. https://doi.org/10.1038/s41586-025-09422-z.

Chicago

Guo, Daya, Dejian Yang, Haowei Zhang, Junxiao Song, Peiyi Wang, Qihao Zhu, Runxin Xu, et al. 2025. “DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning.” Nature 645 (8081): 633–638. https://doi.org/10.1038/s41586-025-09422-z.

Harvard

Guo, D. et al. (2025) ‘DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning’, Nature, 645(8081), pp. 633–638. doi:10.1038/s41586-025-09422-z.

IEEE

D. Guo et al., “DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning,” Nature, vol. 645, no. 8081, pp. 633–638, 2025, doi: 10.1038/s41586-025-09422-z.

BibTeX

@article{guo2025deepseek,
  title     = {{DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning}},
  author    = {Guo, Daya and Yang, Dejian and Zhang, Haowei and Song, Junxiao and Wang, Peiyi and Zhu, Qihao and Xu, Runxin and Zhang, Ruoyu and Ma, Shirong and Bi, Xiao and Zhang, Xiaokang and Yu, Xingkai and Wu, Yu and Wu, Z. F. and Gou, Zhibin and Shao, Zhihong and Li, Zhuoshu and Gao, Ziyi and Liu, Aixin and Xue, Bing and Wang, Bingxuan and Wu, Bowen and Feng, Bei and Lu, Chengda and Zhao, Chenggang and Deng, Chengqi and Ruan, Chong and Dai, Damai and Chen, Deli and Ji, Dongjie and others},
  journal   = {Nature},
  year      = {2025},
  volume    = {645},
  number    = {8081},
  pages     = {633--638},
  doi       = {10.1038/s41586-025-09422-z},
  url       = {https://doi.org/10.1038/s41586-025-09422-z}
}

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

Citations by year

YearCitationsShare of the best year
202011
202211
202377
2024102102
20257,2277227
20269,5659565

Its field-weighted citation impact is 4113.79: it has been cited 4113.79 times as often as the average work of the same field, type and year (1.00 is the world average).

Topic and keywords

Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling

  • DeepSeek-R1
  • large language models
  • reasoning
  • reinforcement learning
  • chain-of-thought prompting
  • self-reflection
  • verification
  • strategic adaptation

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?

Guo, D., Yang, D., Zhang, H., Song, J., Wang, P., Zhu, Q., Xu, R., Zhang, R., Ma, S., Bi, X., Zhang, X., Yu, X., Wu, Y., Wu, Z. F., Gou, Z., Shao, Z., Li, Z., Gao, Z., Liu, A., . . . Zhang, Z. (2025). DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning. Nature, 645(8081), 633–638. https://doi.org/10.1038/s41586-025-09422-z

How many times has this paper been cited?

16,903 times, according to OpenAlex on 15 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: read the free full text at Nature.

Who wrote this paper?

Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Peiyi Wang, Qihao Zhu, Runxin Xu, Ruoyu Zhang, Shirong Ma, Xiao Bi, Xiaokang Zhang, Xingkai Yu and 182 more.

What is the DOI of this paper?

10.1038/s41586-025-09422-z. The permanent link is https://doi.org/10.1038/s41586-025-09422-z.

Bibliographic data from OpenAlex (CC0), last checked 15 November 2026. ResearchGuru does not host this paper; the links go to the publisher or the repository that does.