ComfyUI Gemini Image
A Gemini editing node that turns instructions into image adjustments inside ComfyUI workflows, combining a canvas, references, and inference parameters.
ComfyUI Gemini Edit Image is a custom node for applying text-guided image edits without leaving ComfyUI workflows. The primary image acts as the canvas and up to three references can provide context, while reusable presets and instructions help keep edits consistent.
Alongside the edited image, the node exposes sanitized metrics for latency, retries, requests, tokens, and sampling limits. Edit strength, color and style preservation, temperature, and top_p controls calibrate each change; transient failures receive retries before the result returns to the graph. This lets Gemini fit both direct flows and larger ComfyUI compositions.