An open API service indexing awesome lists of open source software.

https://github.com/openvino-dev-samples/comfyui_openvino


https://github.com/openvino-dev-samples/comfyui_openvino

Last synced: about 1 year ago
JSON representation

Awesome Lists containing this project

README

          

# OpenVINO Node for ComfyUI

This node is designed for optimizing the performance of model inference in ComfyUI by leveraging [Intel OpenVINO toolkits](https://github.com/openvinotoolkit/openvino).

---

- [OpenVINO Node for ComfyUI](#openvino-node-for-comfyui)
- [Supported Hardware](#supported-hardware)
- [Install](#install)
- [Comfy Registry](#comfy-registry)
- [ComfyUI-Manager](#comfyui-manager)
- [Manual](#manual)
- [Instruction](#instruction)

## Supported Hardware

This node can support running model on Intel CPU, GPU and NPU device.You can find more detailed informantion in [OpenVINO System Requirements](https://docs.openvino.ai/2025/about-openvino/release-notes-openvino/system-requirements.html).

## Install

**Prererquisites**

- Install [comfy-cli](https://docs.comfy.org/comfy-cli/getting-started)

The recommended installation method is to use the Comfy Registry.

### Comfy Registry

These nodes can be installed via the [Comfy Registry](https://registry.comfy.org/nodes/comfyui-openvino).

```
comfy node registry-install comfyui-openvino
```

### ComfyUI-Manager

This node can be installed via ComfyUI-Manager in the UI or via the CLI:

```
comfy node install comfyui-openvino
```

### Manual

This node can also be installed manually by copying them into your `custom_nodes` folder and then installing dependencies:

```
cd ComfyUI/custom_nodes
git clone https://github.com/openvino-dev-samples/comfyui_openvino
cd comfyui_openvino
pip install -r requirements.txt
```

## Instruction
To trigger OpenVINO Node for ComfyUI, you can follow the example as reference:
1. Start a ComfyUI server.
```
cd ComfyUI
python3 main.py --cpu --use-pytorch-cross-attention
```
2. Prepare a standard workflow in ComfyUI.

![Step 1](https://github.com/user-attachments/assets/30137084-242b-48ef-8713-fd999168c070)

3. Add OpenVINO Node.

![Step 2](https://github.com/user-attachments/assets/0f9f2841-536b-4e05-8388-49ad219efefd)

4. Connect OpenVINO Node with Model/LoRa Loader.

![Step 3](https://github.com/user-attachments/assets/51d4de0f-c4d2-4e3a-9eb1-3942ef9354ca)

5. Run workflow. Please notice it may need an additional warm-up inference after switching new model.

![Step 4](https://github.com/user-attachments/assets/37a354f2-86eb-4d2a-8ddc-6fc31439ad08)