# Ultralytics

> Computer vision platform from the makers of YOLO for annotating data, training models on cloud GPUs and deploying them.

Ultralytics is the company that develops the YOLO computer vision models and a platform built around them. Ultralytics is used to annotate images and video, train YOLO models on cloud GPUs, and deploy the result to hosted endpoints or export it for edge devices. Work can be done in the browser or through a Python SDK.

**Best for:** Developers and enterprise teams building object detection, segmentation or pose estimation systems who want labelling, training and deployment in one place.

- Official site: https://www.ultralytics.com
- Listed in: [Image Recognition & OCR](https://trendhub.net/ai-tools/category/image-recognition-ocr), [Developer API](https://trendhub.net/ai-tools/category/developer-api)
- Last checked: 2026-10-09 (facts taken from ultralytics.com)
- Source page: https://trendhub.net/ai-tools/ultralytics-yolo-computer-vision-platform

## Key facts

- Platforms: Web, API
- Pricing model: Freemium
- Free plan: Yes
- API: Yes
- Made by: Ultralytics Inc.

## What is Ultralytics?

Ultralytics is the company behind the YOLO family of vision models and a hosted platform that covers the computer vision workflow from raw images to a running model. In the browser, teams label images and video with boxes, polygons, masks, keypoints and oriented boxes, helped by SAM-based annotation that proposes a mask from a click. Training runs on rented cloud GPUs, with live metrics and side-by-side experiment comparison, and can be driven either from a no-code interface or from the Python SDK. Trained models are served from dedicated endpoints in dozens of regions or exported to 22 formats, among them ONNX, TensorRT and CoreML, for edge, mobile and embedded hardware. The models handle detection, segmentation, classification, pose estimation and depth estimation. The platform has a free plan, a Pro plan and an Enterprise plan; software licensing is handled separately.

## Key features

- **Smart Annotation**: Labels images and video across six task types, with SAM proposing masks and with team review and dataset versioning.
- **Cloud Training**: Launches training jobs on a choice of 26 GPU configurations and shows metrics while the run is in progress.
- **YOLO Model Support**: Works natively with YOLOv5, YOLOv8, YOLO11 and YOLO26 in sizes from nano to extra-large.
- **Managed Deployment**: Serves models from dedicated, configurable endpoints with request metrics and logs.
- **Format Export**: Converts a trained model to 22 formats for targets such as NVIDIA Jetson, Raspberry Pi and phones.
- **Python SDK**: Exposes the same training and deployment functions in code for developers who prefer scripts to the web interface.

## How Ultralytics works

1. **Annotate**: Upload images or video and label them with manual tools or SAM-assisted suggestions.
2. **Train**: Pick a YOLO model and a GPU, start the job and watch the metrics.
3. **Deploy or export**: Publish the model to an endpoint in a chosen region, or export it in the format your device needs.

## Use cases

- **Defect Detection on a Production Line**: Train a detector on labelled product photos and run it for quality control.
- **Retail Shelf and Queue Monitoring**: Track stock on shelves and count people waiting using camera feeds.
- **Drone and Aerial Analysis**: Process aerial imagery for agriculture, environmental monitoring and mapping.
- **Robot Perception**: Give an autonomous machine object tracking and navigation cues from its cameras.
- **Sports Tracking**: Follow players and the ball to produce performance analytics.

## Things to know

- **Two license tracks**: Free and Pro plans are labelled AGPL 3.0, while commercial use without open-source obligations requires the Enterprise License.
- **Platform plans and licensing are separate**: The pricing page states that plan fees for the hosted platform do not cover software licensing.
- **GPU time is billed by the hour**: Cloud training draws on credits at hourly GPU rates, and plans include a credit amount.

## FAQs

### What is Ultralytics?

Ultralytics is the developer of the YOLO vision models and of a platform for annotating datasets, training those models on cloud GPUs and deploying them.

### What is Ultralytics used for?

Ultralytics is used to build computer vision applications such as visual inspection, object detection, medical imaging analysis and inventory tracking.

### Is Ultralytics free?

Yes, there is a free plan. Ultralytics says it includes annotation, model training and deployment within set limits, while Pro and Enterprise plans raise those limits.

### Which YOLO versions does Ultralytics support?

The Ultralytics platform supports YOLOv5, YOLOv8, YOLO11 and YOLO26, and the site announces YOLO27 as coming soon.

### Where can Ultralytics models be deployed?

Ultralytics models can be deployed to hosted endpoints in more than 40 regions or exported for edge devices, mobile phones and embedded systems.

### Does Ultralytics have an API or SDK?

Yes. Ultralytics provides a Python SDK and hosted inference endpoints, in addition to its no-code web interface.

### Does Ultralytics offer enterprise licensing?

Yes. Ultralytics has an Enterprise License that removes open-source requirements and comes with dedicated support, custom SLAs and on-premise options.

---

Source: https://trendhub.net/ai-tools/ultralytics-yolo-computer-vision-platform (trendhub)
