> For the complete documentation index, see [llms.txt](https://nx.docs.scailable.net/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://nx.docs.scailable.net/ai-manager-v6.1.4/ai-models-support/importing-models/from-ultralytics.md).

# From Ultralytics

Deploy Ultralytics YOLO models with the AI Manager.

<figure><img src="https://2982163321-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FKGfC0TFSmuMDjbpLErO3%2Fuploads%2Fgit-blob-fd6857997a9be7035d81fd36bc4ca16b40dcf662%2Fultralytics.jpg?alt=media" alt="Ultralytics logo" width="200"><figcaption></figcaption></figure>

Ultralytics develops and maintains popular open-source computer vision libraries including YOLOv8, YOLO11 and YOLO26. These are state-of-the-art object detection models well suited for edge deployment with the AI Manager.

For guides covering other YOLO versions and detectors, see [Common Models](https://nx.docs.scailable.net/ai-manager-v6.1.4/ai-models-support/importing-models/common-models.md).

{% hint style="warning" %}
A plain `yolo export format=onnx` file is not enough. The export has no postprocessing (NMS or confidence filtering), and its inputs and outputs do not use the names the AI Manager expects (see [ONNX requirements](https://nx.docs.scailable.net/ai-manager-v6.1.4/ai-models-support/custom-models/onnx-requirements.md)). Add them with the guide scripts below.
{% endhint %}

## Exporting a YOLO Model for the AI Manager

Use the guide that matches your model, from the [nxai-model-to-onnx](https://github.com/scailable/nxai-model-to-onnx) repository. Each guide exports the model and adds the postprocessing, input and output names the AI Manager needs:

| Model                                       | Guide                                                                                                  |
| ------------------------------------------- | ------------------------------------------------------------------------------------------------------ |
| YOLO26                                      | [`yolo26-to-onnx`](https://github.com/scailable/nxai-model-to-onnx/tree/main/yolo26-to-onnx)           |
| YOLOv12                                     | [`yolov12-to-onnx`](https://github.com/scailable/nxai-model-to-onnx/tree/main/yolov12-to-onnx)         |
| YOLOv10                                     | [`yolov10-to-onnx`](https://github.com/scailable/nxai-model-to-onnx/tree/main/yolov10-to-onnx)         |
| YOLOv8, YOLO11 and other Ultralytics models | [`ultralytics-to-onnx`](https://github.com/scailable/nxai-model-to-onnx/tree/main/ultralytics-to-onnx) |

For a custom or fine-tuned model, call the export script with your checkpoint and input size, then list the classes your model was trained on in `complete_onnx.py`. For example, for YOLO26:

```bash
bash export-to-onnx.sh yolo26n yolo26n.onnx 640
python complete_onnx.py
python test_onnx.py
```

## Upload to AI Manager Cloud

Upload the resulting `.onnx` file to the AI Manager Cloud. The platform automatically converts it for all supported AI accelerators.

{% content-ref url="/pages/Z7ndCO8n1txSBW7CSIq7" %}
[Upload your model](https://nx.docs.scailable.net/ai-manager-v6.1.4/ai-manager-cloud-ui/upload-your-model.md)
{% endcontent-ref %}
