Comfy Console

All the power of ComfyUI. None of the spaghetti.

Free · One HTML file · Runs on your own PC · 100% local & private

ComfyUI is the most powerful open image engine there is — and it looks like a plate of wires. A wall of nodes and noodles that scares off almost everyone who isn’t willing to become an engineer first. That power should not require a diagram.

Comfy Console is a single, calm screen that sits on top of ComfyUI and drives it for you. Every advanced capability is still there — character LoRAs, ControlNet, face and hand repair, detail-adding upscaling, a full editing suite — but each one is a button with a sensible default, not a graph you have to wire. It keeps the ceiling high and drops the floor to the ground.

0
Cloud · Nothing Leaves Your PC
$0
Cost · No Account
Images · No Limits
1
File · One Clean Screen

Why It Exists

Because the gap between “I want to make this” and “I understand this node graph” is where most people give up. Comfy Console closes that gap. It was built to answer a real need: keep every ounce of ComfyUI’s power, but let a person simply type what they want and press one button — the way it always should have worked.

What Makes It Different

It doesn’t dumb ComfyUI down — it dresses it up. Nothing is removed; the hard parts are just handled for you with smart defaults and plain-language controls. And it checks what you have installed, quietly hiding any feature you can’t run, so you only ever see buttons that work. Power underneath, simplicity on top.

Every Pro Feature — One Click Each

Character LoRAs, Done Right

Stack character and style LoRAs, set strength with a slider, and the trigger word appears automatically — click it to drop it in. It reads the word straight from the LoRA file, so you never have to remember it.

Multi-Type ControlNet

Copy a pose, a depth, a line drawing, or an outline from any reference image. Pose, Depth, Lineart, Canny, and Scribble — one model, pick from a dropdown. Turn a sketch into a finished render.

Face, Hand & Foot Repair

The parts AI mangles most, fixed automatically. Detectors find the face, hands, and feet and re-draw just those regions — cleaner every time, in your own style.

Detail-Adding Upscaling

Not just bigger — sharper and more detailed. Tile ControlNet upscaling paints real hair, skin, and fabric detail into the enlarged image instead of just stretching it.

A Full Edit Suite

Drop any image and fix it: paint over a flaw to redraw it (with a soft blend so it never looks pasted on), clone-stamp a blemish away with no AI at all, erase an object so it vanishes, extend the canvas, or clean up a face — with brush, zoom, and undo, no layers to learn.

Train Your Own Characters

Train a character right inside the app — drop in a set of pictures, name it, and the Train tab builds a model of that character for you (one installer, then it lives in the app). Summon the same face in any pose, outfit, or scene. The hard problem in AI art, made routine.

Keep & Explore

Landed a shot you love? One click locks that exact image and generates endless fresh variations of just it — same character and scene, a new roll each time. Turn one lucky result into a whole set.

Quality-Of-Life, Everywhere

Just Type What You Want

Five plain boxes — character, outfit, pose, place, details. No nodes, no wiring. Type words, press Generate. That’s the whole loop.

Tag Shelves & Wildcards

Click real, tested tags from tidy menus — outfits, poses, camera angles, art styles, lighting, weather, colour. Drop any menu in as a {a|b|c} wildcard and every image rolls fresh variety on its own, cycling every option before it repeats.

Recipes & Character Saves

Save your quality settings as a Recipe, and a whole character — prompt plus LoRAs — as a one-click save. Load any character at any quality, instantly.

Forever, Hands-Free

One button generates endlessly on its own. Rotate through your whole cast, randomize the image size, and stack wildcards — so every picture is a fresh character, size, pose, and mood, completely hands-free. Each character files into its own folder.

Smart Defaults

The dialed-in settings that took experts weeks to find are already set — sampler, steps, clip skip, upscaler. It looks great before you touch a thing.

Little Speed Wins

Ctrl+Enter to generate. Ctrl+↑/↓ to strengthen a tag. Reuse a seed to iterate. A PNG-info reader to recover any past image’s settings. The small things add up.

Shows Only What Works

On connect it checks your setup and quietly hides any feature you haven’t installed — so you never see a button that errors. Grow into it, one add-on at a time.

100% Yours

Everything runs on your own machine. No account, no cloud, no cost, no limits, nothing watching. Your images and your words never leave your PC.

Three Steps, Then You’re Making Art

Install ComfyUI Once

Comfy Console drives the free ComfyUI engine. Install it and one image model — a one-time setup, walked through step by step in the guide.

Open the Console

Download one HTML file and open it. It connects to your ComfyUI automatically and checks what you have, showing only the features you can run.

Type & Generate

Describe your picture, press one button, and it appears — on your machine, privately, with no limits. Add the fancy stuff whenever you’re ready.

Full Setup — From Zero How to Use

New to this? The Full Setup walks you through everything — no coding, no Python. Then the guide teaches the buttons.

The power of ComfyUI. The ease it should have had.

Free, private, and unlimited — running entirely on your own PC. Download the Console, connect it to ComfyUI, and start making anything.

Requires a free ComfyUI install. Comfy Console is a single HTML file — nothing to install, nothing that phones home.