This ComfyUI custom node integrates Google's Gemini API, enabling users to leverage advanced AI image generation and manipulation directly within the ComfyUI workflow. It targets ComfyUI users seeking to incorporate powerful, multimodal AI capabilities for creative image synthesis, style transfer, and concept art generation without leaving their preferred environment. The primary benefit is seamless access to Gemini's generative power through a familiar node-based interface.

How It Works

The project provides a dedicated "Gemini 2.0 image" node for ComfyUI. This node acts as an interface to the Google Gemini API, allowing users to send text prompts and optional reference images to generate new visuals. It supports various parameters like model selection, aspect ratio, temperature for controlling creativity, and a random seed for reproducibility. A key feature is its ability to accept multiple reference images, which are used collectively by the API to inform style and content, with the node automatically adjusting the prompt to reflect this multi-image input.

Quick Start & Requirements

• Installation:
  • Manual: Clone the repository into the ComfyUI/custom_nodes directory and install dependencies via requirements.txt using the appropriate Python interpreter (python.exe -m pip install -r requirements.txt).
  • ComfyUI Manager: Search for "Gemini API" within the ComfyUI Manager and install.
• Prerequisites: A working ComfyUI installation and a Google Gemini API key obtained from Google AI Studio.
• Setup: Initial setup requires cloning/installing the node and installing dependencies. The API key needs to be entered once and is then auto-saved.

Highlighted Details

• Supports multi-image input for style reference and content blending.
• Configurable temperature parameter (0.0-2.0) to balance determinism and creativity.
• aspect_ratio selection (free, landscape, portrait, square).
• API key is automatically stored locally after the first entry.

Maintenance & Community

The primary contact and developer is reachable via X (Twitter) at @cychenyue, TikTok at @cychenyue, YouTube at @CY-CHENYUE, BiliBili at @CY-CHENYUE, and 小红书 at @CY-CHENYUE.

Licensing & Compatibility

The license is not explicitly stated in the provided README. Users should verify licensing terms before integrating into commercial or closed-source projects.

Limitations & Caveats

API usage may incur costs or be subject to Google's usage limits. Image generation quality and speed are dependent on Google's server status and network connectivity. Sending reference images to the API has privacy implications. The Gemini API may not always strictly adhere to the requested aspect ratio.