LibreYOLO
MIT-licensed computer vision library. Funds pay for GPU compute: training LibreYOLO26, retraining models on open data, and benchmarks.
Contribute
Become a financial contributor.
Financial Contributions
Top financial contributors
$10 USD since Sep 2026
LibreYOLO is all of us
Our contributors 2
Thank you for supporting LibreYOLO.
Budget
Transparent and open finances.
Credit from Mattia Di Giusto to LibreYOLO •
$8.26 USD
$8.26 USD
--.-- USD
$10.00 USD
About
LibreYOLO is an MIT-licensed computer vision library for detection, segmentation and more, with one train, val and predict API across model families.
Sponsorship pays for work the project can't do without money. It does not pay the maintainer. Funds are loaded as credit onto the project's Vast.ai account, and once a year we publish how much came in and how much was loaded.
What it funds:
- LibreYOLO26, a model trained from scratch and released under MIT, targeting YOLOv8-class accuracy (COCO) and generalization (RF100-VL).
- Retraining model families on open data so their weights can be MIT, starting with YOLO-NAS.
- Training and inference support for vision-language models.
- Benchmarks for visionanalysis.org.
Sponsorship does not buy influence over the roadmap or which PRs are merged. Hardware manufacturers can sponsor by sending a device.
Full program: https://www.libreyolo.com/sponsors
Our team
Xuban Ceccon