I'm a noob so this is what I struggled with and how I fixed it: 1. 'Open in terminal' is not an option when I right-click. Fix: Hold shift, then right-click. 2. Error message when I paste the ComfyUI Manager. Fix: Google "Git" and install it, then try again.
Thank you, I previously used stable diffusion and went through a lot to download each file on stable diffusion which added extra time to make it work but this is a simple process and only took around 15 minutes compared to an hour and a half on stable diffusion.
I installed ComfyUI for the first time because of this video. I'd be happy to see more tutorials like this -- including some that may cover more advanced topics. Thanks for covering this!
I'm glad you installed the ComfyUI Manager, so handy and you can download huggingface models with it. All of the local Stable Diffusion programs are still pretty technical compared to Midjourney or DALL-E 3 but you can control everything about the images. The faceswappers for SD are top notch these days.
Great video. Thanks a lot. I marked this one in my todo list and came back today to go through your process of installing and running it. Great results. I appreciated all the time that you put into this for us.
10 месяцев назад+3
Nice sidenote. you can also drag and drop someons image, if origial in comfy ui and you have the workflow and notes to. And with SDXL you should use 1024 x 1024 as default latent image for better result. Comfy UI is the best creativ tool ever.
Thanks for this! I was just about to return to and update my SDXL Repo but decided to start a fresh instead. It's reassuring to have such an up to date guide. Thankyou 🤜🤛
Matt, INVOKEAI is a manager for SDXL, and is WAY easier to use and install for everyday people with no coding, 1 click install and models come pre-loaded
Everyone's got their favorite UI. I'm liking ComfyUI at the moment but, over time, I plan to play with and create videos around all of them. :) I made this tutorial because ComfyUI specifically was the one I was getting asked about the most. However, if people want the easiest path to using Stable Diffusion, that's why tools like Leonardo and Playground AI exist. Those are way easier than all of this. But some people want to know specifically about things like ComfyUI.
@@mreflowtrue. InvokeAi has by far the easyest to use interface for local installation but under the hood brings also nodes. They are not yet as develloped as comfy but the overal package is agem. I use both. Comfy, but teach invoke to the architects of my company. And for them.. Invoke excels best. You realy have to take a deep look to it. Comfy is for specialists only whyle invoke for all
Comfy UI has a learning curve. But it is the difference between running apps and writing scripts on the command line. Much more powerful and customizable. And the whole workflow is stored in the image itself. There are a lot of web sites that let people share workflows. Just pick the image you like, download it, and drag and drop that image into comfy ui web page. And you can have a dozen comfy ui web pages open at the same time, with different workflows and prompts in each window. You can start by just changing the prompts in other people's workflow, then you can learn how to add a film grain node or a face fix node just before the save node. There are tons of videos here that show you how to build basic workflows, and other videos that show how to use more advanced nodes. :D
By far the easiest walk-through process of ComfyUI installation. I've been wanting to try my hand on this but other tutorials out there are so daunting to follow. Thanks a lot for sharing!
Matt, seeing a US RUclipsr choose Windows is refreshing. No offense to Mac users, but this choice makes you stand out in my eyes even more. Love your channel! 😄
I'm amazed at how you fit in the time to do these tutorials with all the other stuff you have going on. I have not install this but feel like it would be super easy just by following your instructions. Thanks
Great video!! Great tutorial and exactly what I was looking for for a couple of weeks. Nice explaining. For who is having problems with the openpose and it says it's missing a file: My map url was way to long and I shortened it, it gave conflict.
Ah! Amazing!! This is what I’ve been looking for. Gonna be installing this over the weekend for sure. Although these instructional videos aren’t as popular, I’m very grateful that you’re creating them! Thank you!
Is there a way to make it more suitable for weaker pc's? If so, would you be willing to make a little tutorial on it? That would be so awesome!!! Thank you for working this hard! I think this looks really comfy now with the new interface and all.
Thanks Matt. Just getting started with this stuff. Install was a breeze. What is the difference between the base and refiner checkpoints? What is the VAE for? Yes, I will research myself - but a brief mention in the video would be a bonus.
This is a very good tutorial! Nice explanation! I am going to switch to this. You are a absolute journalist and I only follow you. MUCH better than all those old MediaCompany's. PLEASE, stay independent, we need you.
12:52 NOTE: I needed to restart the CMD client in order for the button to appear. Also it's not a bad idea to add "YourDriveLetter:\YourPrefolderIfAny\ComfyUI\python_embeded\Scripts" to PATH.
I'm new to all of this, so quite a bit of what you're talking about it over my head.. can you make a newbie guide to getting into github/huggingface/etc (not a full on github guide or anything, just the things that relate to this stuff) those custom workflow/node graphs are something I really would love to see more on too!! :)
This video is gold!!! I understand why they don't do welll on youtube, but I would love a video that goes even more in dept :) keep up the good job my friend!
Thanks, good video. Honestly I’d like a deeper dive. You did a lot of quick clicking and dragging stuff, and I’m looking at it and thinking “what did you just do and why?”
I need help please, I tried to run the manager on the terminal but it showed this message: git : The term 'git' is not recognized as the name of a cmdlet, function, script file or executable program. Check if you typed the name correctly or, if you included a path, check that the path is correct and try again.
I was just in the process of trying to figure this out and then saw this video. All of your videos are excellent and, of course, this one had my install up and running with no issue (even though I'm running Ubuntu.). Not sure why they are not the most popular but please, keep them comming.
Also - the generated image at the end would have the json data already embedded in it. so you can drag and drop the PNG into the window and it will load the full workflow you used along with the settings.
Amazing video! I had to jump through some extra hoops and figure some stuff out on my own (mainly installing the manager due to the link leading me to the installation of ComfyUI Portable) but it really works!
Like @porcusx I too watched this and followed the steps to install ComfyUI as a non-programmer / non tech person (but interested in seeing how txt to image can work better than Adobe Firefly / DallE etc) I found it very clear and easy to follow compared to others I have recently watched. Well done & thnx Matt!
Now this video answered my Noob question. Question : Do I need to install stable diffusion with ComfyUI ? Answer : No....Stable diffusion is a concept, it is a logic, it is a model for generative AI. 'Checkpoints' , 'VAE', 'Safetensor' are what makes up stable diffusion. So essentially those are collectively stable diffusion while Automatic1111, ComfyUI etc are just front end user interface. If wrong ,please correct my understanding and Sir, it is 'u-lar' not 'oiler'...and Leonhard Euler - he had been the Prince of Mathematics.
Many thanks for the details, would have taken too long for me to dig up all those details from who knows where. More such deepdives would be great. You are a treasure.
Wish I had seen this first, watched a bunch of SD vids and was clueless. This is tge best explanation and ComfyAI is getting installed. Wonder if it will install on my Pi5?
5:19 - You do not need any additional software to unpack 7zip, rar or similar archives on Windows 11 anymore! Just right klick it to unzip it using Window's native functionality ;) Microsoft added support for unpacking these archives some time ago.
Great tutorial ty. For other folks on windows 10 the 'open in terminal' option didn't seem to exist for me so I just navigated there using the commands cd Desktop\whatever. Also once I git cloned comfy manager the refresh button on my browser didn't make it visible, I needed to close the session entirely and re-launch run_nvidia_gpu.bat.
The only problem I have is with the controlnet. No matter what I do it says import/install failed. I googled and didn't find anything related to this so Im hoping somebody here can help me with this xD The rest works perfectly and the video is great, thank you!
Software Engineering: SmythOS’s robust integration capabilities streamline the development of complex software systems, enabling efficient management of large codebases and complex dependencies.
i've tried to add the confyui manager, went to the exact file, clicked the right mouse button, however no option to open in terminal is visible, is there a reason
its because he has installed some special software for the terminal stuff, he forgot to mention that there, but its not a massive problem, you can look at the git page for that manager thing, and it has advice on how to install it, one of them is you download the git software, then you download a particular .bat file and copy paste it to the certain folder in comfyui and then open the .bat file and suddenly it opens git and does all the code and when you restart comfyui the manager thing is there.
1-Go to the folder where you want to open a terminal. 2-Hold the Shift key and then right-click. You should see the "Open PowerShell window here" option.
While comfy is nice, automatic is just better when you need to switch more settings quickly. Or work with an input image with many controlnet that you need to easily change. At least I couldn't get that to run smoothly in comfy. So if you want to use AI to fix images, I recommend automatic.
Fab video. Thank you. I'm looking at using ComfyUi to create comics. If you are able to do more videos, that would be great. around consistency, comic art (maybe). Thank you again.
I use Comfy inside Pinokio instead, but I felt that the image generation time was kind of slow on my M1 Mac. Thanks for the video. 👍 Comfy UI might become my local goto video generation tool in the future. After upgrading to some more AI adapted computer in the future.
For much faster generation install the LCM LoRA and then use the LCM sampler. You can then generate images with only 6 to 8 steps with a CFG Scale of 1.2.
Hey Matt 👋, First off, huge thanks for your video on installing ComfyUI - it's been super helpful! 🙌 I'm really into image generation using fooocus, and I've been curious about automating the upscaling of hundreds of images in a folder to the best possible quality. Your insights on ComfyUI make me think it might be the perfect tool for the job. Could you possibly create more tutorials in this area, specifically on upscaling? I'd love to set up a process that runs overnight on my computer, handling bulk image enhancement. 🌙💻 If you or anyone watching has tips or suggestions on this, I'd be incredibly grateful for your expertise. Looking forward to more of your content! Cheers, Simon 🚀📷
Thanks. Doing videos like this may not provide the highest direct return on you time, but they give you credibility and thus you get an indirect return from them.
THAT, was the easy way to use Stable Diffusion?!! Woah! Not questioning you or ComfyUI. I just can't believe that's the _easy_ way! Would hate to see the real one. As always, thanks for the video Matt. 💯
To get started with Comfy I would suggest Scott Detweilers videos. I resisted comfy until I tried it using his tutorials. Olivio Sarakas also has a couple of good vidoes on Comfy workflows.
Extremely good and highly detailed set of instructions which of course I'm not quite at the point where I can get into that, but I'm so grateful that it's there for when I need to learn it in the future. Now a simple question if anybody wants to answer I'm seeing I have a 4 year old laptop of the I-5 I think it is well then if I were to go to Best buy what can I get that would be able to handle all this and what would the price be.
I am loving the video but when I get to the part where you go to to terminal to right click the ComfyUI manager file, it bugs and says " the cmd github is not recognized as a cmdlet, function, scriptfile, or operable program. Check the spelling of the (before mentioned right click address, but too long to type) ANY help here?
Is it safe to use on real machine as lots of codes are running and beyond the understanding of many real persons? but your video give in depth knowledge to me and that why i subscribed you for more content like this..Keep doing good work..
Amazing guide, thanks Matt :) Don't know if it's just me, but when I have Comfy UI running, Windows 10 refuses to go into Sleep Mode. Anyone else experience this too?
Hi Matt, terrific video as always! Just a tiny heads-up, it's 'cfg scale,' not 'gfc' - easy mix-up though. I'm still a huge fan of A1111. ComfyUI is great if you have a complex workflow which has a lot of steps that need to be changed between renders or if you want to "branch" your workflow to get multiple results. Like a big factory that that has to go trough a lot of building blocks. Inpainting and outpainting is imo best done in A1111 since it is detail work and you have to react to each individual image. Also working with multiple loras makes comfyUI crowded very fast.
@@fintech1378 what do you mean? Let's say i have a single image where i need to fix one small detail and i need a specific lora for that. From the start of both program's until hitting generate / queue prompt how much time / mouseclicks would that take in ComfyUi vs Automatic1111? The answer is: In ComfyUI a lot more! Correct me if I am wrong, but Is there actually a way to do a decent inpaint in ComfyUI apart from installing several custom nodes like ComfyI2I? I stick to my Opinion. For Complex workflows that run fully automated, ComfyUI is a great tool. For Single purpose workflows Automatic1111 is the King.
I'm a noob so this is what I struggled with and how I fixed it:
1. 'Open in terminal' is not an option when I right-click. Fix: Hold shift, then right-click.
2. Error message when I paste the ComfyUI Manager. Fix: Google "Git" and install it, then try again.
I did it. It worked for the isntalation but the MANAGER buttom not comes when I refresh the page.
@@davesousa6723 Just try closing it and opening it again this made it for me.
Wow nice, that's something local that no one can take away from us.
I like your comment
I comment your like @@iresineherb7
Thank you, I previously used stable diffusion and went through a lot to download each file on stable diffusion which added extra time to make it work but this is a simple process and only took around 15 minutes compared to an hour and a half on stable diffusion.
I installed ComfyUI for the first time because of this video. I'd be happy to see more tutorials like this -- including some that may cover more advanced topics. Thanks for covering this!
What kind of specs are you running t on? Does it run well for you so far?
I'm glad you installed the ComfyUI Manager, so handy and you can download huggingface models with it.
All of the local Stable Diffusion programs are still pretty technical compared to Midjourney or DALL-E 3 but you can control everything about the images. The faceswappers for SD are top notch these days.
which face swappers are you talking about can you link them?
great tutorial , please make more detailed tutorial on this , love it super interesting !
Great video. Thanks a lot. I marked this one in my todo list and came back today to go through your process of installing and running it. Great results. I appreciated all the time that you put into this for us.
Nice sidenote. you can also drag and drop someons image, if origial in comfy ui and you have the workflow and notes to. And with SDXL you should use 1024 x 1024 as default latent image for better result. Comfy UI is the best creativ tool ever.
Hey Matt, I did the install on a Mac M1 and it worked perfectly. Its slow but it worked. Thanks for the info !!!
Thanks for this!
I was just about to return to and update my SDXL Repo but decided to start a fresh instead.
It's reassuring to have such an up to date guide.
Thankyou
🤜🤛
Matt, INVOKEAI is a manager for SDXL, and is WAY easier to use and install for everyday people with no coding, 1 click install and models come pre-loaded
Everyone's got their favorite UI. I'm liking ComfyUI at the moment but, over time, I plan to play with and create videos around all of them. :)
I made this tutorial because ComfyUI specifically was the one I was getting asked about the most.
However, if people want the easiest path to using Stable Diffusion, that's why tools like Leonardo and Playground AI exist. Those are way easier than all of this. But some people want to know specifically about things like ComfyUI.
@@mreflow totally makes sense, I'm still gonna download it and try it out, thanks again!
@@mreflowtrue. InvokeAi has by far the easyest to use interface for local installation but under the hood brings also nodes.
They are not yet as develloped as comfy but the overal package is agem. I use both. Comfy, but teach invoke to the architects of my company. And for them.. Invoke excels best. You realy have to take a deep look to it. Comfy is for specialists only whyle invoke for all
Comfy UI has a learning curve. But it is the difference between running apps and writing scripts on the command line. Much more powerful and customizable. And the whole workflow is stored in the image itself. There are a lot of web sites that let people share workflows. Just pick the image you like, download it, and drag and drop that image into comfy ui web page. And you can have a dozen comfy ui web pages open at the same time, with different workflows and prompts in each window. You can start by just changing the prompts in other people's workflow, then you can learn how to add a film grain node or a face fix node just before the save node. There are tons of videos here that show you how to build basic workflows, and other videos that show how to use more advanced nodes. :D
By far the easiest walk-through process of ComfyUI installation. I've been wanting to try my hand on this but other tutorials out there are so daunting to follow. Thanks a lot for sharing!
I love BOTH Introductory ComfyUI (step-by-step) AND advanced ComfyUI! Please ...
Matt, seeing a US RUclipsr choose Windows is refreshing. No offense to Mac users, but this choice makes you stand out in my eyes even more. Love your channel! 😄
I'm amazed at how you fit in the time to do these tutorials with all the other stuff you have going on.
I have not install this but feel like it would be super easy just by following your instructions.
Thanks
Thanks for making this, i know you said it wont get a lot of views. I'm sure it took hours to edit and film for not much ROI, i appreciate it.
Yes, thanks for this. I would love to see you do more on this with animatediff , video to video, prompt travel.
Great tutorial and walk-through! Thank you!
ComfyUI has become my favorite, especially for the ability to maintain workflows easily.
Great video!! Great tutorial and exactly what I was looking for for a couple of weeks. Nice explaining.
For who is having problems with the openpose and it says it's missing a file: My map url was way to long and I shortened it, it gave conflict.
Ah! Amazing!! This is what I’ve been looking for. Gonna be installing this over the weekend for sure. Although these instructional videos aren’t as popular, I’m very grateful that you’re creating them! Thank you!
Is there a way to make it more suitable for weaker pc's? If so, would you be willing to make a little tutorial on it? That would be so awesome!!!
Thank you for working this hard! I think this looks really comfy now with the new interface and all.
Thanks Matt. Just getting started with this stuff. Install was a breeze. What is the difference between the base and refiner checkpoints? What is the VAE for? Yes, I will research myself - but a brief mention in the video would be a bonus.
thanks Matt, was looking for this exact tutorial video made more recently. Detailed explanation as always
This is a very good tutorial!
Nice explanation!
I am going to switch to this. You are a absolute journalist and I only follow you.
MUCH better than all those old MediaCompany's.
PLEASE, stay independent, we need you.
That was fantastic, and if you have the chance to provide more detailed video, it would be truly amazing. Thank you!
12:52 NOTE: I needed to restart the CMD client in order for the button to appear.
Also it's not a bad idea to add "YourDriveLetter:\YourPrefolderIfAny\ComfyUI\python_embeded\Scripts" to PATH.
I'm new to all of this, so quite a bit of what you're talking about it over my head.. can you make a newbie guide to getting into github/huggingface/etc (not a full on github guide or anything, just the things that relate to this stuff) those custom workflow/node graphs are something I really would love to see more on too!! :)
This video is gold!!! I understand why they don't do welll on youtube, but I would love a video that goes even more in dept :) keep up the good job my friend!
Thanks, good video. Honestly I’d like a deeper dive. You did a lot of quick clicking and dragging stuff, and I’m looking at it and thinking “what did you just do and why?”
Thanks Matt appreciate you! Love these deep dives, they're the best on the web. Keep up your legendary work mate!
I need help please, I tried to run the manager on the terminal but it showed this message:
git : The term 'git' is not recognized as the name of a cmdlet, function, script file or executable program. Check
if you typed the name correctly or, if you included a path, check that the path is correct and try again.
Same problem for me
same. any luck?
pls tell me that u have it@@faisalal3435
you have to install the git Extention, i guess, i have the same Problem, Chat.gpt gave me this tip
Edit: it Workt perfectly!
I was just in the process of trying to figure this out and then saw this video. All of your videos are excellent and, of course, this one had my install up and running with no issue (even though I'm running Ubuntu.). Not sure why they are not the most popular but please, keep them comming.
I installed SDiff last month via Discord. But I couldn't get to the right place to use it. Thanks for this video!
Yes, please make more in-depth videos for this app. I have no problem with the length of these videos. Thx for all ur hard work!
If it doesn't show Manager option, just restart the whole thing. Refreshing page did not do it for me.
Also - the generated image at the end would have the json data already embedded in it. so you can drag and drop the PNG into the window and it will load the full workflow you used along with the settings.
Amazing video! I had to jump through some extra hoops and figure some stuff out on my own (mainly installing the manager due to the link leading me to the installation of ComfyUI Portable) but it really works!
Thanks man. I use to draw a lot and I fell out of it. I am thinking of using AI art to get back my artistic spark.
Thank you so much for this information and details, we need to know more about ComfyUI especially how to make videos with it.
I understand the time and effort It takes to give away a video like this. Thank you, congratullations. You are the BEST!
I've just installed my new 3090!! and was looking exactly for a tutorial for Comfy UI. Many thanks Matt, perfect timing!! XD
Like @porcusx I too watched this and followed the steps to install ComfyUI as a non-programmer / non tech person (but interested in seeing how txt to image can work better than Adobe Firefly / DallE etc) I found it very clear and easy to follow compared to others I have recently watched. Well done & thnx Matt!
Now this video answered my Noob question. Question : Do I need to install stable diffusion with ComfyUI ? Answer : No....Stable diffusion is a concept, it is a logic, it is a model for generative AI. 'Checkpoints' , 'VAE', 'Safetensor' are what makes up stable diffusion. So essentially those are collectively stable diffusion while Automatic1111, ComfyUI etc are just front end user interface.
If wrong ,please correct my understanding and Sir, it is 'u-lar' not 'oiler'...and Leonhard Euler - he had been the Prince of Mathematics.
Many thanks for the details, would have taken too long for me to dig up all those details from who knows where. More such deepdives would be great. You are a treasure.
Thanks, King! Can you do an updated video on your latest top 5 favorite models for under 32 gigs of RAM??? That would be AMAZING!
Just walked into this, this video made it clear on how to use comfy - thanks
Thank you so much for this tutorial man. Really in depth and comprehensive for people that are just starting to get into the mix like me. Much love
Thanks Matt, I really appreciate the Comfy Ui install video shout out. Great video as always, thanks. I hope you've recovered from CES as well.👍
Wish I had seen this first, watched a bunch of SD vids and was clueless. This is tge best explanation and ComfyAI is getting installed. Wonder if it will install on my Pi5?
5:19 - You do not need any additional software to unpack 7zip, rar or similar archives on Windows 11 anymore! Just right klick it to unzip it using Window's native functionality ;)
Microsoft added support for unpacking these archives some time ago.
Maybe its the weed that I smoked but this was quite the tutorial. Amazing stuff. All those tweaking abillities ! (and uncencored)
Great tutorial ty. For other folks on windows 10 the 'open in terminal' option didn't seem to exist for me so I just navigated there using the commands cd Desktop\whatever. Also once I git cloned comfy manager the refresh button on my browser didn't make it visible, I needed to close the session entirely and re-launch run_nvidia_gpu.bat.
The only problem I have is with the controlnet. No matter what I do it says import/install failed. I googled and didn't find anything related to this so Im hoping somebody here can help me with this xD
The rest works perfectly and the video is great, thank you!
Software Engineering: SmythOS’s robust integration capabilities streamline the development of complex software systems, enabling efficient management of large codebases and complex dependencies.
i've tried to add the confyui manager, went to the exact file, clicked the right mouse button, however no option to open in terminal is visible, is there a reason
its because he has installed some special software for the terminal stuff, he forgot to mention that there, but its not a massive problem, you can look at the git page for that manager thing, and it has advice on how to install it, one of them is you download the git software, then you download a particular .bat file and copy paste it to the certain folder in comfyui and then open the .bat file and suddenly it opens git and does all the code and when you restart comfyui the manager thing is there.
same^^^
1-Go to the folder where you want to open a terminal.
2-Hold the Shift key and then right-click. You should see the "Open PowerShell window here" option.
While comfy is nice, automatic is just better when you need to switch more settings quickly. Or work with an input image with many controlnet that you need to easily change. At least I couldn't get that to run smoothly in comfy. So if you want to use AI to fix images, I recommend automatic.
Fab video. Thank you. I'm looking at using ComfyUi to create comics. If you are able to do more videos, that would be great. around consistency, comic art (maybe). Thank you again.
Good stuff i always wanted to run my own node but lacked the programmer skills
I use Comfy inside Pinokio instead, but I felt that the image generation time was kind of slow on my M1 Mac. Thanks for the video. 👍 Comfy UI might become my local goto video generation tool in the future. After upgrading to some more AI adapted computer in the future.
For much faster generation install the LCM LoRA and then use the LCM sampler. You can then generate images with only 6 to 8 steps with a CFG Scale of 1.2.
@@ScottLahteine Thanks will check that out. 😀
Hey Matt 👋,
First off, huge thanks for your video on installing ComfyUI - it's been super helpful! 🙌 I'm really into image generation using fooocus, and I've been curious about automating the upscaling of hundreds of images in a folder to the best possible quality. Your insights on ComfyUI make me think it might be the perfect tool for the job.
Could you possibly create more tutorials in this area, specifically on upscaling? I'd love to set up a process that runs overnight on my computer, handling bulk image enhancement. 🌙💻
If you or anyone watching has tips or suggestions on this, I'd be incredibly grateful for your expertise. Looking forward to more of your content!
Cheers,
Simon 🚀📷
Thanks. Doing videos like this may not provide the highest direct return on you time, but they give you credibility and thus you get an indirect return from them.
This is the most helpful video on the internet by far. Hope he make more advance video for comfyui
Thanks for the updated video. please keep tutorials coming! 😁
THAT, was the easy way to use Stable Diffusion?!! Woah! Not questioning you or ComfyUI. I just can't believe that's the _easy_ way! Would hate to see the real one.
As always, thanks for the video Matt. 💯
yeah, it's quite technical and requires IQ > 120
hope poor artists can adapt, may be big pharmas invent some smartifying pills for them or smth
@@VincentPride1986 I hope they do too . 👍
make sure you specifically get python version 3.10.6 or else you will get errors running the bat file
Try Debate Assistant! It helps you back up arguments with facts and stats.
That Slack notification made me wonder where the hell I got that notification, only to find out it's your Slack, not mine :D
To get started with Comfy I would suggest Scott Detweilers videos. I resisted comfy until I tried it using his tutorials. Olivio Sarakas also has a couple of good vidoes on Comfy workflows.
First time watching a video about Stable Diffusion, awesome explanation!! thank you!
yes, more tutorials on custom AI please
Thanks
great step by step tutorial, your explanation is perfect, easy to follow, thanks bro !!
Thankyou for this Matt,I was able to follow along and install sucessfully.
The tools change so fast that I'm glad you make theses videos
Extremely good and highly detailed set of instructions which of course I'm not quite at the point where I can get into that, but I'm so grateful that it's there for when I need to learn it in the future.
Now a simple question if anybody wants to answer I'm seeing I have a 4 year old laptop of the I-5 I think it is well then if I were to go to Best buy what can I get that would be able to handle all this and what would the price be.
It didnt work with the manager-Add-On. My terminal says, that it dont know what "git" means. I made it exactly like the video
same problem
Same here
Do you have git installed and know what it is.
Google "git" and install it for Windows
Matt you are the great! I love your detailed tutorial; I'm going to try it now... Thanks!
Thank you so much for this clear and actually usefull step by step tutorials 🙏🙏🙏
Been on A111 forever but maybe it's time to give Comfy a try
I am loving the video but when I get to the part where you go to to terminal to right click the ComfyUI manager file, it bugs and says " the cmd github is not recognized as a cmdlet, function, scriptfile, or operable program. Check the spelling of the (before mentioned right click address, but too long to type) ANY help here?
same
yah same issue here, anyone know the solution
same
Is it safe to use on real machine as lots of codes are running and beyond the understanding of many real persons? but your video give in depth knowledge to me and that why i subscribed you for more content like this..Keep doing good work..
Great vid! Yes more Cumfy and Stable Diffusion and local install or Colab install please
Couldn't add manager. I don't see "Open in terminal" when I right click.
Great clip, thx for great content. I have discovered a lot of great tools because of you, keep doing !
When you do it daily it's not hard... But for some reason I had the problems with XL models, so whatching your video
super awesome instruction on installation. much appreciated matt !
Thank you for putting this together, it was super useful.
Matt: watched but its alot of time for what our kcxu radio station needs but will share with the youngers
Amazing guide, thanks Matt :)
Don't know if it's just me, but when I have Comfy UI running, Windows 10 refuses to go into Sleep Mode. Anyone else experience this too?
This is amazing!!!!! Thanks so much this de mystified everything for me!!!!! Keep up the great work
Great stuff, thanks Matt! More in depth tutorials please 💻
Excellent tutorial! Up and running. Thank you
Easy to follow and solid tutorial. I hope you'll eventually create advanced tutorials on the many nodes available for ComfyUI. Great job!
Really great video. I like your tutorials a lot. I learned quite a bit from you. Thank you for doing these videos.
Wow that was great I look forward to trying it out. Thanks😊
Excellent tutorial. Thanks for taking the time. I hope people are clicking on the thumbs up.
Thank you - this was one of the best tutorials I've come across for a long time.
Maybe the real *_comfy user interface_* is all the friends we made along the way.
Thankyou for taking the time to post this! Very helpful!
Hi Matt, terrific video as always! Just a tiny heads-up, it's 'cfg scale,' not 'gfc' - easy mix-up though. I'm still a huge fan of A1111. ComfyUI is great if you have a complex workflow which has a lot of steps that need to be changed between renders or if you want to "branch" your workflow to get multiple results. Like a big factory that that has to go trough a lot of building blocks. Inpainting and outpainting is imo best done in A1111 since it is detail work and you have to react to each individual image. Also working with multiple loras makes comfyUI crowded very fast.
Whats the benefit of a1111 compared to comfyui again
@@fintech1378 what do you mean?
Let's say i have a single image where i need to fix one small detail and i need a specific lora for that. From the start of both program's until hitting generate / queue prompt how much time / mouseclicks would that take in ComfyUi vs Automatic1111?
The answer is: In ComfyUI a lot more!
Correct me if I am wrong, but Is there actually a way to do a decent inpaint in ComfyUI apart from installing several custom nodes like ComfyI2I?
I stick to my Opinion. For Complex workflows that run fully automated, ComfyUI is a great tool.
For Single purpose workflows Automatic1111 is the King.
@@AIKnowledge2Go ok i think comfyui has one node that can just do that thanks