Tune in to YOLO Vision 2025!
September 25, 2025
10:00 — 18:00 BST
Hybrid event
Yolo Vision 2024

Ultralytics YOLO models

Power vision AI solutions with Ultralytics YOLO models, enabling fast
and accurate real-time analysis of images and videos through efficient computer vision tasks.

Trusted by industry leaders

Shell LogoNIST LogoNGIA LogoMenlo Security LogoArgonne LogoMusashi LogoDIA Logo
Shell LogoNIST LogoNGIA LogoMenlo Security LogoArgonne LogoMusashi LogoDIA Logo
Shell LogoNIST LogoNGIA LogoMenlo Security LogoArgonne LogoMusashi LogoDIA Logo

Try our models

Explore how Ultralytics YOLO models works directly in your browser.

0.00
0.00
Running inference...
Something went wrong...
Upload Image
Queue of passengersGiraffe and zebras at the NC zoo
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

Our models’ impact

Streamline processes across industries with our cutting-edge vision AI models.
Speed, accuracy and ease-of-use powered by Ultralytics.

GitHub stars

110k+

Model downloads

120M+

Ultralytics YOLO usages

2B+

Open-source contributors

900+

The evolution of YOLO models

2015
YOLO
Developed by Joseph Redmon and team.
2016
YOLOv2
Created by Joseph Redmon and Ali Farhadi.
2018
YOLOv3
Developed by Joseph Redmon and Ali Farhadi.
2018
YOLOv4
Developed by Alexey Bochkovskiy and team.
2020 - Ultralytics YOLOv5
Ultralytics YOLOv5
It redefined object detection with fast, accurate performance and a user-friendly design. Developed by Ultralytics, YOLOv5 became the industry standard thanks to its simple PyTorch implementation and reliable real-time results.
Learn more
2022
YOLOv6
Introduced by Meituan and team.
2022
YOLOv7
Released by Chien-Yao Wang and team.
2023 - Ultralytics YOLOv8
Ultralytics  YOLOv8
Created by Ultralytics, this version expanded support to new tasks like instance segmentation, classification, and pose estimation, while improving performance across the board in a clean, modular format.
Learn more
2024
YOLOv9
Developed by Chien-Yao Wang and team.
2024
YOLOv10
Designed by Ao Wang for Tsinghua University.
2024 - Ultralytics YOLO11
Ultralytics YOLO11
Just as easy to use as Ultralytics YOLOv8, Ultralytics released YOLO11, a model that supports the computer vision tasks users are familiar with while offering better accuracy, precision, and speed.
Learn more
2025 - Ultralytics YOLO26
Ultralytics YOLO26
The latest model from Ultralytics, YOLO26, is better, faster and smaller. Built for real-world deployment, it delivers unmatched efficiency and accuracy.
Learn more

Ultralytics YOLO models at a glance

Ultralytics YOLOv5
Ultralytics YOLOv8
Ultralytics YOLO11
Speed
Image processing time
1.06 ms
0.99 ms
1.05 ms
Accuracy
mAP50-95*
34.3%
37.3%
39.5%
Supported tasks
Object detection
Ultralytics
YOLOv5
Ultralytics
YOLOv8
Ultralytics
YOLO11
Image classification
Ultralytics
YOLOv5
Ultralytics
YOLOv8
Ultralytics
YOLO11
Instance segmentation
Ultralytics
YOLOv5
Ultralytics
YOLOv8
Ultralytics
YOLO11
Object tracking
Ultralytics
YOLOv5
Ultralytics
YOLOv8
Ultralytics
YOLO11
Pose estimation
Ultralytics
YOLOv5
Ultralytics
YOLOv8
Ultralytics
YOLO11
OBB detection
Ultralytics
YOLOv5
Ultralytics
YOLOv8
Ultralytics
YOLO11

Why choose Ultralytics YOLO?

Optimized for edge & cloud deployment

Optimized for edge & cloud deployment

Ultralytics YOLO models are designed for edge & cloud deployment, bringing flexibility to your real-time vision AI projects. This means, faster and reliable results across infrastructures.

1

Seamless integration
with built-in security

Seamless integration
with built-in security

Easily integrate Ultralytics YOLO models across platforms, devices and for all your vision AI workflows. Your data stays local and you can be confident that it remains protected.

2

Clear licensing and compliance

Clear licensing and compliance

We offer flexible licensing options for our Ultralytics YOLO models, including academic, open-source, and enterprise licenses. Choose the option that best fits your needs.

3

Backed by a global community

Backed by a global community

At Ultralytics, we pride ourselves on our global and active community. With dedicated support channels, forums, and regular updates, you'll always have the resources you need to succeed.

4

State-of-the-art accuracy and efficiency

State-of-the-art accuracy and efficiency

We strive to make our models better, pushing the boundaries of vision AI. Model performance is at the core of our R&D led by accuracy, efficiency and streamlined infrastructure.

5

See our models in action

Learn more

Frequently asked questions

What are the different Ultralytics YOLO models?

Our models include Ultralytics YOLOv5, Ultralytics YOLOv8, and Ultralytics YOLO11. Each version enhances accuracy, speed, and supports multiple vision tasks such as object detection, segmentation, and pose estimation for versatile AI applications.

Which YOLO model is best?

The best Ultralytics YOLO model for your use-case depends on your needs. The latest Ultralytics YOLO11 model offers the latest improvements in accuracy, speed, and multi-task support.

What is Ultralytics in Python?

The Ultralytics Python package provides easy access to Ultralytics YOLO models, making it possible for developers to train, validate, and deploy computer vision tasks like object detection and segmentation within Python projects.

What are the different variants of Ultralytics YOLO11?

The different Ultralytics YOLO11 variants are YOLO11n, YOLO11s, YOLO11m, YOLO11l, and YOLO11x. They come in different sizes to offer a balance between speed and accuracy for various uses.

Get in touch

Get started