Elamur Research⌕
☰
home / papers / 2025 / HunyuanImage 3.0 Technical Report

HunyuanImage 3.0 Technical Report

Paper · 2025 · Tencent Hunyuan Foundation Model Team

Summary

Most top-tier image generators, such as Seedream 4.0, Nano Banana and GPT-Image, are closed. That makes them hard to study or reproduce. Tencent's Hunyuan team responds with HunyuanImage 3.0, an openly released model that handles both image understanding and image generation inside one autoregressive system.

The team started from the Mixture-of-Experts LLM Hunyuan-A13B and added a pre-trained vision encoder and a VAE. Projection layers map both into the LLM's embedding space. Generation uses diffusion over VAE latents, following the Transfusion and JanusFlow approach. A small reasoning dataset teaches the model to plan and rewrite a prompt before drawing, which is a form of native Chain-of-Thought.

Training relied on a heavily filtered corpus with bilingual, multi-level captions checked by agents. The released checkpoint covers only text-to-image generation.

In side-by-side human judgments, the model beats its predecessor and roughly matches leading commercial systems. This gives researchers a strong open foundation for multimodal experimentation.

Key points

  • Scale: The model has more than 80B total parameters, with about 13B active per token. The authors call it the largest open-source image generator so far.
  • Data:
    • A three-stage pipeline cut a pool of over 10B raw images to roughly 5B training images, keeping less than 45%.
    • Filtering removed watermarks, AIGC and low-quality content, and scored images for clarity and aesthetics.
    • A separate set of more than 100M image pairs and multi-image samples was mined from image clusters and video.
  • Captions:
    • Captions are bilingual (English/Chinese) and follow a hierarchical schema, ranging from about 30 to 1,000 words.
    • An OCR Agent and a Named Entity (IP) Agent support the captioner.
    • A bidirectional verification step keeps only captions that agree with the agents' outputs.
  • Evaluation:
    • On the MLLM-based SSAE benchmark, the model matches leading systems in every field.
    • In a GSB study (1,000 prompts, 100+ professional raters), its relative win rates were:
      • +14.10% over HunyuanImage 2.1
      • +1.17% over Seedream 4.0
      • +2.64% over Nano Banana
      • +5.00% over GPT-Image
  • Analysis: Experts in the MoE layers become more specialized by modality (image vs. text) in deeper layers. Image-to-image capability is still in training.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Fields, topics and concepts

📚 Shelves 2The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksComputer Vision and Pattern Recognition · 2,869 works

Cite

Cite as Tencent Hunyuan Foundation Model Team. 2025. HunyuanImage 3.0 Technical Report. arXiv:2509.23951.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.48550/arxiv.2509.23951 · arXiv 2509.23951

@misc{team2025hunyuanimage,
  title = {{HunyuanImage 3.0 Technical Report}},
  author = {Tencent Hunyuan Foundation Model Team},
  year = {2025},
  doi = {10.48550/arxiv.2509.23951},
  eprint = {2509.23951},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2509.23951}
}

References

45 references, 17 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffusion probabilistic models. Advances in neural information processing systems, 33:6840–6851, 2020. find
  2. Jiaming Song, Chenlin Meng, and Stefano Ermon. Denoising diffusion implicit models. arXiv preprint arXiv:2010.02502, 2020.
  3. Prafulla Dhariwal and Alexander Nichol. Diffusion models beat gans on image synthesis. Advances in neural information processing systems, 34:8780–8794, 2021. find
  4. Yang Song, Jascha Sohl-Dickstein, Diederik P Kingma, Abhishek Kumar, Stefano Ermon, and Ben Poole. Score-based generative modeling through stochastic differential equations. arXiv preprint arXiv:2011.13456, 2020.
  5. Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Björn Ommer. High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10684–10695, 2022. find
  6. Tero Karras, Miika Aittala, Timo Aila, and Samuli Laine. Elucidating the design space of diffusion-based generative models. Advances in neural information processing systems, 35: 26565–26577, 2022. find
  7. Yaron Lipman, Ricky TQ Chen, Heli Ben-Hamu, Maximilian Nickel, and Matt Le. Flow matching for generative modeling. arXiv preprint arXiv:2210.02747, 2022.
  8. Xingchao Liu, Chengyue Gong, and Qiang Liu. Flow straight and fast: Learning to generate and transfer data with rectified flow. arXiv preprint arXiv:2209.03003, 2022.
  9. Zijian Zhang, Zhou Zhao, and Zhijie Lin. Unsupervised representation learning from pre-trained diffusion probabilistic models. Advances in neural information processing systems, 35: 22117–22130, 2022. find
  10. Zijian Zhang, Zhou Zhao, Jun Yu, and Qi Tian. Shiftddpms: Exploring conditional diffusion models by shifting diffusion trajectories. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 37, pages 3552–3560, 2023. find
  11. Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. Advances in neural information processing systems, 30, 2017.
  12. William Peebles and Saining Xie. Scalable diffusion models with transformers. In Proceedings of the IEEE/CVF international conference on computer vision, pages 4195–4205, 2023.
  13. Team Seedream, Yunpeng Chen, Yu Gao, Lixue Gong, Meng Guo, Qiushan Guo, Zhiyao Guo, Xiaoxia Hou, Weilin Huang, Yixuan Huang, Xiaowen Jian, Huafeng Kuang, Zhichao Lai, Fanshi Li, Liang Li, Xiaochen Lian, Chao Liao, Liyang Liu, Wei Liu, Yanzuo Lu, Zhengxiong Luo, Tongtong Ou, Guang Shi, Yichun Shi, Sh… — [arXiv:2509.20427](https://arxiv.org/abs/2509.20427) find
  14. Google. Nano banana, 2025. URL https://developers.googleblog.com/en/introdu cing-gemini-2-5-flash-image. ↗
  15. OpenAI. Gpt-image, 2025. URL https://platform.openai.com/docs/models/gpt-i mage-1. ↗
  16. Chenfei Wu, Jiahao Li, Jingren Zhou, Junyang Lin, Kaiyuan Gao, Kun Yan, Sheng ming Yin, Shuai Bai, Xiao Xu, Yilei Chen, Yuxiang Chen, Zecheng Tang, Zekai Zhang, Zhengyi Wang, An Yang, Bowen Yu, Chen Cheng, Dayiheng Liu, Deqing Li, Hang Zhang, Hao Meng, Hu Wei, Jingyuan Ni, Kai Chen, Kuan Cao, Liang… — Qwen-Image Technical Report
  17. Tencent Hunyuan. Hunyuanimage 2.1, 2025. URL https://github.com/Tencent-Hunyu an/HunyuanImage-2.1. ↗
  18. Tencent Hunyuan Team. Hunyuan-a13b. https://github.com/Tencent-Hunyuan/Hunyu an-A13B, 2024. ↗
  19. Jean-Baptiste Alayrac, Jeff Donahue, Pauline Luc, Antoine Miech, Iain Barr, Yana Hasson, Karel Lenc, Arthur Mensch, Katherine Millican, Malcolm Reynolds, et al. Flamingo: a visual language model for few-shot learning. Advances in neural information processing systems, 35: 23716–23736, 2022.
  20. Haotian Liu, Chunyuan Li, Qingyang Wu, and Yong Jae Lee. Visual instruction tuning. Advances in neural information processing systems, 36:34892–34916, 2023.
  21. Haotian Liu, Chunyuan Li, Yuheng Li, and Yong Jae Lee. Improved baselines with visual instruction tuning. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 26296–26306, 2024.
  22. Chunting Zhou, Lili Yu, Arun Babu, Kushal Tirumala, Michihiro Yasunaga, Leonid Shamis, Jacob Kahn, Xuezhe Ma, Luke Zettlemoyer, and Omer Levy. Transfusion: Predict the next token and diffuse images with one multi-modal model. arXiv preprint arXiv:2408.11039, 2024. find
  23. Yiyang Ma, Xingchao Liu, Xiaokang Chen, Wen Liu, Chengyue Wu, Zhiyu Wu, Zizheng Pan, Zhenda Xie, Haowei Zhang, Xingkai Yu, et al. Janusflow: Harmonizing autoregression and rectified flow for unified multimodal understanding and generation. In Proceedings of the Computer Vision and Pattern Recognitio…
  24. Yunpeng Luo, Junlong Du, Ke Yan, and Shouhong Ding. Lareˆ 2: Latent reconstruction error based method for diffusion-generated image detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 17006–17015, 2024. find
  25. Ziyin Zhou, Yunpeng Luo, Yuanchen Wu, Ke Sun, Jiayi Ji, Ke Yan, Shouhong Ding, Xiaoshuai Sun, Yunsheng Wu, and Rongrong Ji. Aigi-holmes: Towards explainable and generalizable ai-generated image detection via multimodal large language models. arXiv preprint arXiv:2507.02664, 2025. — [arXiv:2507.02664](https://arxiv.org/abs/2507.02664) find
  26. Ruoxin Chen, Junwei Xi, Zhiyuan Yan, Ke-Yue Zhang, Shuang Wu, Jingyi Xie, Xu Chen, Lei Xu, Isabel Guan, Taiping Yao, et al. Dual data alignment makes ai-generated image detector easier generalizable. arXiv preprint arXiv:2505.14359, 2025. — [arXiv:2505.14359](https://arxiv.org/abs/2505.14359) find
  27. Tinglei Feng, Yingjie Zhai, Jufeng Yang, Jie Liang, Deng-Ping Fan, Jing Zhang, Ling Shao, and Dacheng Tao. Ic9600: A benchmark dataset for automatic image complexity assessment. IEEE Transactions on Pattern Analysis and Machine Intelligence, (01):1–17, 2023. doi: 10.1109/TPAMI.2022.3232328. — [doi:10.1109/tpami.2022.3232328](https://doi.org/10.1109/tpami.2022.3232328) find
  28. William Peebles and Saining Xie. Scalable diffusion models with transformers. In Proceedings of the IEEE/CVF international conference on computer vision, pages 4195–4205, 2023.
  29. Patrick Esser, Sumith Kulal, Andreas Blattmann, Rahim Entezari, Jonas Müller, Harry Saini, Yam Levi, Dominik Lorenz, Axel Sauer, Frederic Boesel, Dustin Podell, Tim Dockhorn, Zion English, and Robin Rombach. Scaling rectified flow transformers for high-resolution image synthesis. In ICML, 2024. URL… find
  30. Zhimin Li, Jianwei Zhang, Qin Lin, Jiangfeng Xiong, Yanxin Long, Xinchi Deng, Yingfang Zhang, Xingchao Liu, Minbin Huang, Zedong Xiao, Dayou Chen, Jiajun He, Jiahao Li, Wenyue Li, Chen Zhang, Rongwei Quan, Jianxiang Lu, Jiabin Huang, Xiaoyan Yuan, Xiaoxiao Zheng, Yixuan Li, Jihong Zhang, Chao Zhang,… — Hunyuan-DiT: A Powerful Multi-Resolution Diffusion Transformer with Fine-Grained Chinese Understanding
  31. Black Forest Labs. Flux, 2024. URL https://blackforestlabs.ai/. ↗
  32. Weijie Kong, Qi Tian, Zijian Zhang, Rox Min, Zuozhuo Dai, Jin Zhou, Jiangfeng Xiong, Xin Li, Bo Wu, Jianwei Zhang, Kathrina Wu, Qin Lin, Junkun Yuan, Yanxin Long, Aladdin Wang, Andong Wang, Changlin Li, Duojun Huang, Fang Yang, Hao Tan, Hongmei Wang, Jacob Song, Jiawang Bai, Jianbing Wu, Jinbao Xue,… — HunyuanVideo: A Systematic Framework For Large Video Generative Models
  33. Jinheng Xie, Weijia Mao, Zechen Bai, David Junhao Zhang, Weihao Wang, Kevin Qinghong Lin, Yuchao Gu, Zhijie Chen, Zhenheng Yang, and Mike Zheng Shou. Show-o: One single transformer to unify multimodal understanding and generation. arXiv preprint arXiv:2408.12528, 2024. — [arXiv:2408.12528](https://arxiv.org/abs/2408.12528) find
  34. Xiaokang Chen, Zhiyu Wu, Xingchao Liu, Zizheng Pan, Wen Liu, Zhenda Xie, Xingkai Yu, and Chong Ruan. Janus-pro: Unified multimodal understanding and generation with data and model scaling. arXiv preprint arXiv:2501.17811, 2025.
  35. Chaorui Deng, Deyao Zhu, Kunchang Li, Chenhui Gou, Feng Li, Zeyu Wang, Shu Zhong, Weihao Yu, Xiaonan Nie, Ziang Song, et al. Emerging properties in unified multimodal pretraining. arXiv preprint arXiv:2505.14683, 2025. — [arXiv:2505.14683](https://arxiv.org/abs/2505.14683) find
  36. Chao Liao, Liyang Liu, Xun Wang, Zhengxiong Luo, Xinyu Zhang, Wenliang Zhao, Jie Wu, Liang Li, Zhi Tian, and Weilin Huang. Mogao: An omni foundation model for interleaved multi-modal generation. arXiv preprint arXiv:2505.05472, 2025. — [arXiv:2505.05472](https://arxiv.org/abs/2505.05472) find
  37. Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016.
  38. Jianlin Su, Murtadha Ahmed, Yu Lu, Shengfeng Pan, Wen Bo, and Yunfeng Liu. Roformer: Enhanced transformer with rotary position embedding. Neurocomputing, 568:127063, 2024.
  39. Bram Wallace, Meihua Dang, Rafael Rafailov, Linqi Zhou, Aaron Lou, Senthil Purushwalkam, Stefano Ermon, Caiming Xiong, Shafiq Joty, and Nikhil Naik. Diffusion model alignment using direct preference optimization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, p… find
  40. Junzhe Li, Yutao Cui, Tao Huang, Yinping Ma, Chun Fan, Miles Yang, and Zhao Zhong. Mixgrpo: Unlocking flow-based grpo efficiency with mixed ode-sde. arXiv preprint arXiv:2507.21802, 2025. — [arXiv:2507.21802](https://arxiv.org/abs/2507.21802) find
  41. Xiangwei Shen, Zhimin Li, Zhantao Yang, Shiyi Zhang, Yingfang Zhang, Donghao Li, Chunyu Wang, Qinglin Lu, and Yansong Tang. Directly aligning the full diffusion trajectory with fine-grained human preference. arXiv preprint arXiv:2509.06942, 2025. — [arXiv:2509.06942](https://arxiv.org/abs/2509.06942) find
  42. Zhengyang Geng, Mingyang Deng, Xingjian Bai, J Zico Kolter, and Kaiming He. Mean flows for one-step generative modeling. arXiv preprint arXiv:2505.13447, 2025. find
  43. Kaiyi Huang, Kaiyue Sun, Enze Xie, Zhenguo Li, and Xihui Liu. T2i-compbench: A comprehensive benchmark for open-world compositional text-to-image generation. Advances in Neural Information Processing Systems, 36:78723–78747, 2023. find
  44. Dhruba Ghosh, Hannaneh Hajishirzi, and Ludwig Schmidt. Geneval: An object-focused framework for evaluating text-to-image alignment. Advances in Neural Information Processing Systems, 36:52132–52152, 2023. find
  45. Yuang Peng, Yuxin Cui, Haomiao Tang, Zekun Qi, Runpei Dong, Jing Bai, Chunrui Han, Zheng Ge, Xiangyu Zhang, and Shu-Tao Xia. Dreambench++: A human-aligned benchmark for personalized image generation. arXiv preprint arXiv:2406.16855, 2024. — [arXiv:2406.16855](https://arxiv.org/abs/2406.16855) find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/team-2025-hunyuan-hunyuanimage-3-0-technical/