Ultralytics YOLO27: Join the waitlist.
License YOLO
Ultralytics YOLO

Ultralytics YOLO, the leader in computer vision

Production-proven computer vision, relied on by millions of developers and leading enterprises. License it for your own stack, or build on Ultralytics Platform.

Trusted by the world's leading organizations

The standard for every vision task. Seven tasks, one model family, from cloud to edge.

Object detection: metal castings on a conveyor belt, with bounding boxes labeled part_damaged and crack.
Instance segmentation: panoramic dental X-ray with teeth segmented by individual masks and numbered tooth labels.
Semantic segmentation: highway scene colored by class, vehicles and trees in green, road in blue, roadsides in pink.
Depth estimation: warehouse workers in hard hats as a depth map, nearest people in warm colors, background in blue.
Image classification: yellow package on a curved conveyor belt, classified as box.
Pose estimation: warehouse worker in a hard hat and safety vest carrying a box on one shoulder, with body keypoints joined into a skeleton.
Oriented object detection (OBB): aerial view of solar panels on a dark shingled roof, each detected with a rotated box labeled panel.
Object detection: metal castings on a conveyor belt, with bounding boxes labeled part_damaged and crack.
Instance segmentation: panoramic dental X-ray with teeth segmented by individual masks and numbered tooth labels.
Semantic segmentation: highway scene colored by class, vehicles and trees in green, road in blue, roadsides in pink.
Depth estimation: warehouse workers in hard hats as a depth map, nearest people in warm colors, background in blue.
Image classification: yellow package on a curved conveyor belt, classified as box.
Pose estimation: warehouse worker in a hard hat and safety vest carrying a box on one shoulder, with body keypoints joined into a skeleton.
Oriented object detection (OBB): aerial view of solar panels on a dark shingled roof, each detected with a rotated box labeled panel.
Object detection: metal castings on a conveyor belt, with bounding boxes labeled part_damaged and crack.
Instance segmentation: panoramic dental X-ray with teeth segmented by individual masks and numbered tooth labels.
Semantic segmentation: highway scene colored by class, vehicles and trees in green, road in blue, roadsides in pink.
Depth estimation: warehouse workers in hard hats as a depth map, nearest people in warm colors, background in blue.
Image classification: yellow package on a curved conveyor belt, classified as box.
Pose estimation: warehouse worker in a hard hat and safety vest carrying a box on one shoulder, with body keypoints joined into a skeleton.
Oriented object detection (OBB): aerial view of solar panels on a dark shingled roof, each detected with a rotated box labeled panel.

Our models' impact

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

138.8K+
GitHubGitHub stars
340.8M+
Downloads
4.1B+
Ultralytics YOLO usages / day
1.2K+
Open-source contributors
Ultralytics YOLO licensing

Integrate YOLO into systems you already have

Ship Ultralytics YOLO in your own solutions, with an Enterprise License for commercial use.

  • Commercial use: No AGPL-3.0 obligations for closed source.
  • Your infrastructure: On-premise, edge, or your own cloud.
  • 22 export formats: ONNX, TensorRT, CoreML, and more.
  • Dedicated support: Custom SLAs and direct engineer access.
Ultralytics logo connected to Annotate, Train, and Deploy.
Ultralytics Platform

Annotate, train, and deploy in one place

Label data, train on cloud GPUs, and deploy globally, with no infrastructure needed.

Our customers

Start with open models, add enterprise terms

The models developers build on

Ultralytics YOLO is the default starting point for real-time detection across 7 tasks, from research to production.

Python code that loads a YOLO26 model and ships it to production, commented with Ultralytics Enterprise License.

License it for production

Remove open-source obligations, with dedicated support, custom SLAs, and on-premise deployment.

Security & compliance

Enterprise-grade security, certified

Independently certified and audited, so your data stays protected in transit and at rest.

SOC 2 Type I

Attested against AICPA Security, Availability, and Confidentiality TSC

ISO 27001:2022

Certified information security management system

Encryption

Data encrypted in transit and at rest, across all storage and transport

GDPR & CCPA

Aligned with global data protection regulations

Where Ultralytics YOLO delivers results

Computer vision in logistics

Fewer supply chain errors

From the warehouse floor to the loading dock, Ultralytics YOLO automates counting, sorting, and inspection across every step of the supply chain.

  • Warehouse & inventory: Automate counts across every aisle.
  • Parcel & baggage: Detect and spot damage at conveyor speed.
  • Transport maintenance: Catch faults on vehicles early.
  • Loading docks: Verify loads and track vehicle dwell times.
  • License plate recognition: Automate gate access and yard management.
Move goods faster with Ultralytics YOLO

Frequently asked questions

  • Free use, including experimentation and evaluation, requires releasing your entire project's source code under AGPL-3.0, including models you train. For closed-source evaluation or deployment, request an Ultralytics Enterprise License.

  • YOLO runs on CPUs, NVIDIA GPUs, Jetson and Raspberry Pi. The Python package trains and runs models on your own hardware without collecting your images. Platform deploys to 42 global regions, and Enterprise offers on-premises data and compute options.

  • Install the open-source package with pip install ultralytics and follow the training guide, or use Platform to upload images, label them with smart annotation and train on cloud GPUs.

  • For new projects, start with YOLO26. YOLO11, YOLOv8 and YOLOv5 remain supported for existing workflows. Sizes range from nano to extra-large: smaller models run faster on edge devices, larger ones can be more accurate. Compare models.

  • The Free plan has no subscription fee; cloud training and deployments use credits. Free and Pro plans carry AGPL-3.0 open-source obligations, so only Enterprise includes a commercial license. Compare plans.

  • Ultralytics builds Ultralytics YOLO, real-time computer vision models used by millions of developers, and Ultralytics Platform, where you annotate data, train on cloud GPUs and deploy. Use YOLO through the open-source Python package, or the whole workflow in your browser with Platform.

  • Each license is quoted for your organization's size and use case, and covers current and future YOLO models released during the license term. Request a quote.

Build on the models the industry already uses

Talk to our team about licensing Ultralytics YOLO for your products, or start on Ultralytics Platform.