ComfyUI-Copilot Documentation

When connected to LMStudio, ComfyUI-Copilot supports the following capabilities:

• Model listing from LMStudio
• Chat completions
• Streaming responses
• All core ComfyUI-Copilot features (including workflow generation and debugging)

The Parameter Debug Interface is a modular component system within ComfyUI Copilot designed for parameter debugging. It is organized into several functional directories:

• Core Components: ParameterDebugInterfaceV2.tsx (enhanced wrapper) and ParameterDebugInterfaceNew.tsx (original component).
• Types: types/parameterDebugTypes.ts contains common type definitions.
• Utilities: Specialized logic is split into utils/ including localStorageUtils.ts, textInputUtils.ts, parameterUtils.ts, interfaceUtils.ts, and searchUtils.ts.
• UI Elements: screens/ and modals/ contain the various view layers, while styles/highlightPulseStyle.ts manages CSS animations.
• Exports: All components and utilities are accessible via index.ts.

After installing and running ComfyUI, follow these steps to activate the service:

1. Start Service: Locate and click the Copilot activation button on the left panel of the ComfyUI interface.
2. Generate API Key: Click the key icon button, enter your email address in the popup, and wait for the API Key to be sent to your email. Paste the received key into the input field and click Save.
3. Configure Custom Models: If you wish to use your own OpenAI models, click the settings icon to configure the dialogue model and the workflow generation model.

Important Notice: Due to service adjustments, built-in node information queries, workflow recommendations, and workflow generation services may be unavailable. To use Agent capabilities, you must provide your own API Key and Base URL in the settings page.

The recommended installation method is to clone the repository directly into your ComfyUI/custom_nodes directory and install the requirements. This is generally more reliable than using ComfyUI Manager.

For Linux/macOS/Standard Python:

1. Navigate to ComfyUI/custom_nodes.
2. Clone the repository.
3. Install requirements via pip.

For Windows (using ComfyUI embedded Python): Use the python_embedded\python.exe executable to ensure dependencies are installed into the correct environment.

# Standard installation
cd ComfyUI/custom_nodes
git clone https://github.com/AIDC-AI/ComfyUI-Copilot.git
cd ComfyUI-Copilot
pip install -r requirements.txt

# Windows embedded Python installation
python_embedded\python.exe -m pip install -r requirements.txt

To use Copilot, you must obtain and save an API Key:

1. In the Copilot interface, click the "*" button.
2. Enter your email address; the system will send the API Key to you.
3. Paste the key into the input field and click Save to complete activation.

After installing and running ComfyUI, follow these steps to activate the service:

1. Launch Service: Click the Copilot activation button on the left side of the ComfyUI panel.
2. Generate API Key: Click the * button in the Copilot interface. Enter your email address in the popup. An API Key will be sent to your email.
3. Enter API Key: Paste the received API Key into the input box and click save to activate.
4. Configure Models: Click the * button to separately configure the chat model and the workflow generate model (e.g., using OpenAI or LMStudio).

Important Service Notice: The default API service for Node information query, Job recommendations, and Workflow generation has been suspended. To use Agent-related capabilities, you must enter your own API Key and Base URL in the Settings page.

You can install the extension using the ComfyUI Manager:

1. Open ComfyUI Manager.
2. Click on Custom Nodes Manager.
3. Search for ComfyUI-Copilot.
4. Click the install button.

Note:

• It is recommended to run ComfyUI with sudo python main.py to ensure the Manager has necessary permissions.
• If updates fail, try deleting the folder or performing an uninstall/reinstall.
• The Git installation method is generally more stable than the Manager method.

To use local language models via LMStudio, configure the LLM/API settings in the ComfyUI-Copilot interface.

Configuration Requirements:

• API Key: Leave this completely empty. LMStudio does not require an API key.
• Server URL: Enter the LMStudio server address. You must append /v1 to the end of the URL.

Steps:

1. Open ComfyUI-Copilot in your browser.
2. Click the settings/configuration button.
3. Locate the "LLM Configuration" or "API Configuration" section.
4. Enter the Server URL (e.g., http://localhost:1234/v1).
5. Click "Verify" to test the connection.
6. Click "Save".

To ensure your LMStudio integration is working correctly, follow these testing steps:

1. Start LMStudio server with a model loaded.
2. Configure ComfyUI-Copilot with your LMStudio URL.
3. Test connection using the verify button in the UI.
4. Load models: Verify that the models loaded in LMStudio appear in the ComfyUI-Copilot model dropdown.
5. Test chat functionality: Send a simple query to ensure the LLM responds.
6. Test features: Verify that ComfyUI-Copilot features like workflow generation and debugging function using the local model.