What’s the practical difference between Low, Medium, and High strength?

Low is conservative—minimal risk of halos and detail hallucination; ideal for portraits and lightly compressed photos. Medium balances clarity and fidelity, suitable as a default for mixed content. High is aggressive—great for small text, logos, and heavily compressed images, but can introduce ringing or over-sharpened textures on some photos.

How should I evaluate the A/B comparison from ImageCompare?

Inspect at 100–200% zoom. Check skin, hair, foliage, text edges, and repetitive patterns. Favor results that improve clarity without halos, zippering, waxy skin, or moiré. If artifacts appear, try a lower strength.

Can the model invent details that weren’t in the original?

Yes—Wonder 3.5 is generative and may synthesize fine detail, especially at High strength. This can be desirable for rescuing low-quality sources but may reduce documentary fidelity. Choose a lower strength when accurate texture is more important than perceived sharpness.

Do I need any preprocessing like denoising before enhancement?

Usually no. TopazImageEnhanceV2 already targets noise and clarity. If your source has extreme compression artifacts or heavy sensor noise, start with Low or Medium to avoid amplifying defects, and only move to High if text/graphics still need more bite.

Image Upscale: Topaz Wonder 3.5

This workflow compares the original image against a Topaz Wonder 3.5–enhanced result to help you pick the right enhancement strength for your needs. Using LoadImage to ingest your source, TopazImageEnhanceV2 applies generative upscaling with three selectable strengths—Low, Medium, and High—targeting fine detail recovery, noise reduction, and crisper text/graphics. ImageCompare then outputs a side-by-side composite so you can visually assess changes at a glance, while SaveImageAdvanced saves both the enhanced image and the comparison view.

The key trade-offs lie in how aggressively Wonder 3.5 invents or amplifies detail. Low is conservative and best when you want to preserve natural textures (e.g., skin) with minimal risk of halos. Medium balances clarity and fidelity for most photos. High is most assertive and can dramatically sharpen small text, logos, and structured graphics or rescue heavily compressed images—but it’s also where over-sharpening and hallucinated detail are most likely. By rerunning the same image with different strength settings and reviewing each A/B comparison, you can quickly converge on the best setting for your subject matter.

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.

9bcda9e23e19.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

Frequently Asked Questions