Hello,
I am interested in using the YOLO NAS model for both research and development as well as commercial purposes. I have noticed that NVIDIA recently acquired Deci.ai, the company behind the YOLO NAS model. Additionally, it appears that the GitHub repository for SuperGradients, which previously hosted the YOLO NAS model, is now deprecated.
Could anyone provide information on whether there is a specific license required to use the YOLO NAS model for commercial purposes, or if it is available for free use? Any details regarding the terms and conditions of usage would be greatly appreciated.
Thank you for your assistance.
Best regards,
Hi @ProfessorHT,
Thank you for reaching out.
Please note that NVIDIA and Deci do not support Yolo-NAS, as the product is no longer maintained. While the open-source version remains available on GitHub, it is free for non-commercial use only. Unfortunately, commercial use is not permitted, and we are no longer issuing licenses.
Let us know if you have any other questions.
Hi @sophwats,
Thank you for your response. Just to clarify, even if I train the model from scratch, would I still be unable to use it for commercial purposes?
Best Regards
Hi @ProfessorHT8
The YOLO-NAS architecture is licensed under Apache, allowing you to train it from scratch and use it commercially.
Best,
sophie
Hi,
I’d like to clarify the commercial licensing situation for YOLO-NAS.
In this post, it is stated that YOLO-NAS is free for commercial use under Apache 2.0, as long as it’s trained from scratch without pretrained weights.
However, the official LICENSE.YOLONAS.md on GitHub says “The weights and architecture of the YOLO-NAS model are licensed under the Deci License – for research and non-commercial use only.”
Can you confirm:
- Does this restriction still apply to the architecture, even if no pretrained weights are used?
- Is there an officially supported commercial path for using YOLO-NAS?
Thanks in advance for clarifying this!