Welcome to your guide on running Replicate models in ComfyUI! In this article, we will walk you through setting up your environment, installing the necessary components, and utilizing the custom nodes to make the most out of the Replicate models. Let’s dive in!

Setting Up Your Environment

Before you can start running Replicate models, it’s essential to set your Replicate API token in your environment. This token allows you to authorize your requests and interact with the Replicate API effectively.

Get Your API Token

To get your API tokens, visit Replicate API Tokens. It’s advisable to create a new token for secure access.

Set the API Token in Your Environment

To pass in your API token when running ComfyUI, you can use the following commands:

• On MacOS or Linux:

export REPLICATE_API_TOKEN=r8_************; python main.py

• On Windows:

set REPLICATE_API_TOKEN=r8_************; python main.py

Installing ComfyUI Custom Nodes

Now that your API token is set, let’s move on to installing the custom nodes necessary for running Replicate models.

Direct Installation Steps

• Navigate to the ComfyUI directory:

cd ComfyUI

• Clone the repository:

git clone https://github.com/replicate/comfyui-replicate

• Enter the directory:

cd comfyui-replicate

• Install the requirements:

pip install -r requirements.txt

Supported Replicate Models

You can check out the supported models by viewing the `supported_models.json` file. This file contains a list of models packaged by default for your convenience.

Updating and Adding Models

Keeping your models updated is crucial for optimal performance. Here’s how to do that:

• Run the following command to update all model nodes:

.import_schemas.py

• To add models, ensure that they only return simple text or image outputs. For audio or video outputs, there may be limitations.
• Follow these steps to add a model:
• Add the model to `supported_models.json` (e.g., `fofrconsistent-character`).
• Run `.import_schemas.py` to import the new model and update all schemas.
• Restart ComfyUI.
• Use the model in your workflow; it will be titled ‘Replicate [model author model name]’.

Roadmap for Improvements

There are exciting plans on the horizon for this custom node package! Some areas under investigation include:

• Support for more types of Replicate models (audio and video first).
• Implementing logs, prediction status, and progress tracking via tqdm.

Troubleshooting

If you encounter issues while setting up or using the Replicate models in ComfyUI, here are some tips to help you out:

• Ensure that your API token is set correctly in your environment.
• Verify that you have all the required Python packages installed.
• Check the console output for error messages that might guide you to the issue.
• For more insights, updates, or to collaborate on AI development projects, stay connected with fxis.ai.

Conclusion

At fxis.ai, we believe that such advancements are crucial for the future of AI, as they enable more comprehensive and effective solutions. Our team is continually exploring new methodologies to push the envelope in artificial intelligence, ensuring that our clients benefit from the latest technological innovations.

By following the steps outlined above, you can harness the full potential of Replicate models within ComfyUI. Happy coding!