ComfyUI Beginner Tutorial — Step-by-Step Guide & Transcript

[00:00] In this video, I’ll show you how to use ComfyUI  step by step. ComfyUI is one of the most   popular free and open-source tools for creating  AI-generated content on your own computer. People  

[00:12] use it to generate images, create videos, enhance  photos, and build powerful AI workflows using   a visual node-based interface. In this tutorial,  we'll start from the very beginning. We'll install   ComfyUI, set up an image model, generate our first  AI image, and then learn how workflows and nodes  

[00:29] work so you can start building your own projects.  I'm Kevin, and let's dive in. To get started,   head to the following website. You'll find a link  right here at the bottom of the screen. Here on   the homepage, you'll notice that ComfyUI offers  both a desktop version and cloud version. For this  

[00:46] tutorial, we'll use the desktop version since it's  free to use, runs directly on your own computer,   and gives you full access to your models and  workflows. The cloud version can be a good option   if you don't have a powerful computer, or maybe  you'd rather not manage the hardware yourself,  

[01:01] although there is a cost associated with it. Let's  click on download and then download again. Once   the installer finishes downloading, run through  the install process. After installation completes,  

[01:13] launch ComfyUI, and you'll see the following  screen. For this tutorial, we'll select local,   so everything runs directly on your own  computer. Down below, I'll agree to the terms,   and then let's click on continue. Next, let's  give this installation a name. I'll set mine  

[01:27] to ComfyUI. For most people, you can leave the  remaining settings as is, and then down below,   let's click on continue. Once ComfyUI launches,  you'll land on the main getting started screen.   Now, right away, you can see that ComfyUI supports  much more than just the image generation. Over  

[01:42] here on the left-hand side, you'll find categories  for images. Here we have videos, audio, 3D models,   large language models, and more. To help you  get up and running quickly, let's go back to the  

[01:56] getting started section, and let's start with text  to image. I'll click here. ComfyUI now loads a   complete image generation workflow. Now, notice in  the top right-hand corner that ComfyUI is telling   us that we're missing a required model. Think of  ComfyUI as the application that allows us to build  

[02:13] and run AI workflows. The model itself is what  actually creates the image. To see what we need,   let's click on show missing models. Right here,  ComfyUI shows us the exact model required by this  

[02:26] workflow. To install the model, right over here,  let's click on download all. Right up on top,   we can now check on the download progress.  Now that the model has finished downloading,   let's generate our first image. Let's take a  quick look at the workflow. You can zoom in  

[02:39] and out using your mouse wheel, and you can press  the middle mouse button to pan around the canvas.   You can also use the mini map in the bottom  right-hand corner to quickly navigate around   larger workflows. Now, notice that this workflow  is divided into several different sections. Right  

[02:56] over here, we start by loading the model that  we just downloaded. Next, we enter a prompt,   and this describes the image that we'd like to  create. Now, I want to generate an image of a   chocolate chip cookie, so right over here,  I'll enter in the following prompt. Now,  

[03:11] below that we have the negative prompt. This  tells the model what we'd like to avoid in the   generated image. So, over here, I'll enter  in my text. If you want to follow along,   you'll find these same exact prompts in the  description of this video. Next, we can configure  

[03:25] the image size. Here we could set the width, we  could also set the height, and also the number of   images we'd like to generate. Now I'll leave the  default settings in place since they work well   with this model. Next, let's go up and we'll look  at the sampling section. This is where ComfyUI  

[03:40] generates the image. You'll notice several  different settings here. Now, as a beginner,   I recommend leaving these at their default  values. Now there is one setting worth knowing,   and that's the seed. You'll find that right here  at the top of the list. Changing the seed allows  

[03:55] you to generate different variations of the same  image all while using the same prompt. For now,   let's leave all of these settings as is. Then, in  the top right-hand corner, let's now click on Run.   And there we have our first AI generated image.  It also automatically saved the image to your  

[04:11] computer. Now, you might be wondering how do you  view all of your generated images or assets? Over   on the left-hand side, we have this button for  assets. When I hover over, you'll also notice that   there is an associated shortcut key. You can press  the letter A. When I click on that, here I can see  

[04:26] all the different images that I've generated.  If I right click, over here, I could save it   elsewhere on my computer or down below, I could  also delete the asset if I'd like. Right above,   I'll close this view. Now that we've successfully  generated an image, let's take a closer look at  

[04:40] how this workflow works and how we could modify  it to produce a different result. Now, right   now we're generating just a chocolate chip cookie.  Let's see what happens when we make a few changes.   So right over here, we have the prompt field,  and let's update this text to instead create,  

[04:55] let's say, a double chocolate chip cookie with a  melted chocolate center. That sounds so good. Now,   if we scroll down just a little bit, here  we have the image size. And previously,   we went with a 512 pixel by 512 pixel image, and  that produced a square image. But let's actually  

[05:11] try a different aspect ratio. So right over  here, let's change the width to, let's say 768. I'll enter in that value. And then right up on  top, let's try running this. I'll click here.  

[05:24] And there we have a double chocolate chip cookie,  and now it's also a wider image. Now that we've   generated a few different images, let's take a  closer look at how this workflow is organized.   One of the things that makes ComfyUI unique  is that the workflows are built using nodes.  

[05:40] Each node performs a specific task. Here's an  example. We have our first node, and that loads   the checkpoint or the model that we downloaded.  And together, all of these different nodes create   the workflow that generates this final image. Now,  as you look across this canvas, notice how all of  

[05:58] these different nodes are connected together with  all of these different lines. Data flows from one   node to the next until we eventually arrive at  the finished image. Now, the nice thing about   this approach is that we're not limited to the  workflow exactly as it exists today. We could  

[06:14] easily extend it by adding new nodes. I'll zoom  in right here on this last stage. To add a node,   the option to add a node. I'll click on that, and  that opens up a menu with hundreds of different  

[06:30] nodes that you can use to customize your workflow.  Now, I'd like to add an image upscaling node. So   here within the image at the very top of the list,  we have the option for upscaling. And right here,   we can upscale the image by. I'll click on that.  Right over here, it's now added a new upscaling  

[06:45] node. And let's try upscaling the image, let's say  by 2X. Here I'll enter 2 and click on OK. However,   right now, this node is not connected to the  overall workflow. So over here, let's take the  

[06:57] output of the decoding node, and here I'll feed  it into the upscaling step. And over here, I'll   take the output of the upscale step, and let's  feed that into the save image step. That way,   we'll upscale the image before we end up saving  it. Now, let's run the workflow again. In the top  

[07:12] right, I'll click on run. And right there, you  can see that we've now enhanced the image simply   by adding a single node to the workflow. Right  down below, you'll now see the new dimensions   of this image. We've effectively upscaled it  by 2. This is the core idea behind ComfyUI. By  

[07:27] combining and connecting different nodes, you can  build workflows that generate images, edit photos,   create videos, process audio, and much more.  Now that we've customized this workflow, let's   save it so we can use it again later. Right up on  top, you'll see all of your different workflows.  

[07:42] I'll right click on this one, and right here,  we see the option to save. I'll click on this.   Right here, we can enter in a file name. I'll call  mine the upscale image generator, and down below,   let's click on confirm. To get back to workflow,  over on the left-hand side, click on workflows,  

[07:56] and there you can see the workflow that we just  saved. You can also share workflows with other   people, which is one of the reasons ComfyUI has  such a large and active community. Instead of   building everything from scratch, you can start  from an existing workflow and modify to fit your  

[08:11] exact needs. Here, I'll close out of workflows.  Before we wrap up, let's take one more look at   One of the things I like about ComfyUI is  that you have a large collection of starter  

[08:23] templates to get you up and running quickly. Now,  in addition to text to image, which we just used,   here you'll find other fun ones, like image to  video. Here, you have image editing. Here we   have image to video again, image to 3D. You even  have text to song, so lots of different options  

[08:38] and fun things that you can play with. Now, I  recommend clicking in, downloading the models, and   then exploring how they're built. Every workflow  here is made up of nodes, and you could use these   templates as a starting point and then customize  them to fit your own needs. And that's how to use  

[08:53] ComfyUI. Now that you understand the basics, go  ahead and experiment with the built-in templates   and start building workflows of your own. If you'd  like to learn more about AI tools and how to get   the most out of them, consider subscribing. Thanks  for watching, and I'll see you in the next video.