Wish your favorite images could grow larger without losing their quality? Upscaling with ComfyUI makes it easy to enlarge pictures while keeping them clear and sharp, perfect for both personal projects and professional work.

In this guide, we'll show you how to use ComfyUI for upscaling, so you can enjoy stunning, vibrant images every time.

Verified to work on ThinkDiffusion Build: April 1, 2025

Why do we specify the build date? ComfyUI and custom node versions that are part of this workflow that are updated after this date may change the behavior or outputs of the workflow.

How to Upscale an Image in ComfyUI

Installation guide

💡
Download the workflow and drag & drop it into your ComfyUI window, whether locally or on ThinkDiffusion. If you're using ThinkDiffusion, it's necessary to use at minimum the Turbo 24gb machine, but we do recommend the Ultra 48gb machine.

Custom Nodes

If there are red nodes in the workflow, it means that the workflow lacks the required nodes. Install the custom nodes in order for the workflow to work.

  1. Go to ComfyUI Manager  > Click Install Missing Custom Nodes
  1. Check the list below if there's a list of custom nodes that needs to be installed and click the install.

Models

For this guide you'll need 5 models.

💡
Download the necessary models (flux1-dev.safetensors, t5xxl_fp16.safetensors, clip_l.safetensors, ae.safetensors and 4xNomos8kDAT.pth) using the ComfyUI Model Manager.
  1. Go to ComfyUI Manager  > Click Model Manager
  1. When you find the exact model that you're looking for, click install and make sure to press refresh when you are finished.

Model Path Source

Use the model path source if you prefer to install the models using model's link address and paste into ThinkDiffusion MyFiles using upload URL.

Model Name Model Link Address ThinkDiffusion Upload Directory
flux1-dev.safetensors
📋 Copy Path
.../comfyui/models/diffusion_models/
t5xxl_fp16.safetensors
📋 Copy Path
.../comfyui/models/clip/
clip_l.safetensors
📋 Copy Path
.../comfyui/models/clip/
ae.safetensors
📋 Copy Path
.../comfyui/models/vae/
4xNomos8kDAT.pth
📋 Copy Path
.../comfyui/models/upscale_models/

Step-by-step Workflow Guide

Steps Recommended Nodes
1. Load an Image ThinkDiffusion-StableDiffusion-ComfyUI-Image-Upscaler-load-image.png
2. Set the Models ThinkDiffusion-StableDiffusion-ComfyUI-Image-Upscaler-set-models.png
3. Set the Upscale Settings

Follow the upscale settings as seen on the image.

The upscale_by serves as a multiplying factor for the upscale. If set to 2, it will double in size.

The workflow uses a flux-based model and the recommended tile width and tile height is 1024 x 1024
ThinkDiffusion-StableDiffusion-ComfyUI-Image-Upscaler-set-upscale-settings.png
4. Check the Image Preview ThinkDiffusion-StableDiffusion-ComfyUI-Image-Upscaler-check-image-preview.png

Examples


Any of these workflows will work on your local machine. If you’re having issues with installation or slow hardware, you can try any of these workflows on a more powerful GPU in your browser with ThinkDiffusion.

If you enjoy ComfyUI and you want to test out different workflows, then feel free to check out these 10 cool ComfyUI workflows here.

Or to have your hand at video, check out these workflows:

Discover why Wan 2.1 is the best AI video model right now.
0:00 /0:05 1× The recently released Wan 2.1 is a groundbreaking open-source AI video model. Renowned for its ability to exceed the performance of other open-source models like Hunyuan and LTX, as well as numerous commercial alternatives, Wan 2.1 delivers truly incredible text2video and image2video generations
AI Video Speed: How LTX is Reshaping Video2Video as We Know It
LTX Video to Video can deliver really powerful results in amazing speed and that’s just the power of LTX - speed. With this video2video workflow we’ll be transforming input videos in AI counterparts with amazing efficiency.
Enhanced Video Generation with Hunyuan and LoRA
We now can use LoRAs together with the AI video model Hunyuan. Why? To keep character or object consistency in a video.