What exactly is being compared in this workflow?

The workflow compares the original image (from LoadImage) with the edited image generated by GeminiNanoBanana2V2. ImageCompare places them side-by-side for visual assessment; no numeric metrics are computed.

How do I ensure a fair, readable comparison?

Keep the source image constant and change only the edit instruction or a single parameter at a time. Use consistent output size, and review the same regions (faces, edges, text) for drift or artifacts.

When should I choose Nano Banana 2 Lite for edits?

Choose it when you need fast, coherent edits to an existing image—such as color tweaks, lighting adjustments, or minor object-level changes—where turnaround speed is important.

Why does the edit sometimes alter areas I didn’t mention?

Under-specified instructions can lead to broader changes. Make your intent explicit (what to change and what to keep), and iterate with more targeted guidance to reduce unintended drift.

Nano Banana 2 Lite: Image Edit

This comparison workflow shows a clean, side-by-side view of how Nano Banana 2 Lite (via the GeminiNanoBanana2V2 node) transforms a source image based on your edit instruction. The pipeline loads your original image, applies the edit with the model, and then places the original and result next to each other using ImageCompare so you can quickly judge what changed and what stayed consistent.

Methodologically, the comparison is controlled: both panes derive from the exact same source image (LoadImage), and only the model-driven edit differs. This makes it easy to evaluate trade-offs such as speed versus fidelity, the extent of stylistic drift, and the preservation of structure or identity. Expect Nano Banana 2 Lite to prioritize responsiveness and efficient edits; in practice, that can mean fast, coherent changes for common tasks (color tweaks, small object additions/removals), while very intricate or large, global transformations may require tighter instructions or iterative passes.

API

Use this workflow from code

Every Comfy workflow is a JSON graph. The payload below is this workflow, exactly as ComfyUI runs it — fetch it from the URL, keep it in version control, or load it in ComfyUI and run it node by node.

62cb2b168265.json

Fetching workflow JSON…

Run it from TypeScript or Python with the Comfy SDK. The same code targets Comfy Cloud or a ComfyUI you host yourself — only the base URL changes.

// Install (beta)
npm i @comfyorg/sdk

// Run this workflow (TypeScript)
import { Comfy } from "@comfyorg/sdk";

const client = new Comfy({ apiKey: "comfyui-..." });
const wf = await client.workflows.fromFile("workflow_api.json");
const job = await client.run(wf);
await job.getOutputs("<output-node-id>")[0].toFile("output.png");

The SDK takes a workflow in API format: open this workflow in ComfyUI and use File → Export Workflow (API).

Comfy Cloud API access requires a plan with an API key.

SDK docs Get an API key

FAQ