Download ComfyUI Studio

DigitAll

Now the v328 10 rar is the latest studio. No need to update. Do not replace. Install in a new folder. Replace the models from the old to the new studio

Hugh Milstein

I looked at the update YouTube video but I can't find the link for the update.rar file. Or should I just download all 10 rar files and replace older version of comfy directory. I don't have anything configured yet

DigitAll

Read this post.

hitesh parmar

how do i downlad the package?

Stephen Orsatti

RTX 3080 10gb

DigitAll

Yes

Laurent Cresson

Rtx 3090?

red baron

and for runpod?

DigitAll

not for videos. only for images

jonh doe

rtx 5060?

Laurent Cresson

are you sure?

DigitAll

Just was a maintenance. Disconnect discrod in Patreon settings and connect again.

Hans

What happened to the discord server ??????

DigitAll

Hi. Tomorrow

Laurent Cresson

Hi when does complete installation files v305 Will be downloadable?

DigitAll

For image generation yes. For video no

ai.playhardgopro

Is any point for me to download this version if I have RTX3060 12GB ?

DigitAll

Yes every week will be more.

Andres Lentijo

Update 288 All models download size actually is 512GB

DigitAll

do you use discord?

Laurent Cresson

what's going on with version 279? I downloaded and installed it before noticing that it had disappeared from patreon Any bugs? the missing loras with XXX I just hope I won't be annoyed when the next upgrade comes along.

Kara Concepcion

Nevermind, it was a Patreon glitch! For anyone else who subscribes and can't join the server, leave your existing servers on the Patreon backend (Settings > Apps) and disconnect Discord, then reconnect Discord and join the server.

Kara Concepcion

Hi, I just subscribed and I cannot find an active Discord invite?

DigitAll

Yes i se this issue. You have no access to discord? Write me on discord i will add you to the group

john Greene

I paid for a one-year membership, and now it's saying I'm on free status. What's that about, sir?

Serdar Polat

Hi Bro discord blocked in Turkey can you send me message pls

Laurent Cresson

Hi dose anyonecan tell me where is flux kontext in studio?

Jose M. Aranda

Thank you for adding Flux's Kontext Dev. I have a request, if possible: could you add the 8-step Hyper Flux LoRA and Tea cache for Kontext Dev? I've seen some YouTube videos where they've included them in Flux Kontext workflows, and they significantly help accelerate image rendering. Could you also implement uscaling and a genre viewer as you have in other workflows?. Thank you. On the other hand there is a problem with images and the image size node. When you upload images with another size you get this error; regardless of the size of the image, it always fails: The size of tensor a (11974) must match the size of tensor b (8160) at non-singleton dimension 1. Even if you put the same dimensions of the image in the node, this error occurs.

DigitAll

Go to discord "download studio" channel

Greg Waters

I have joined....linked to discord....still don't see the download link...

DigitAll

Ok will send you in message

Ulaş Aksan

Hi discord blocked in Turkey can you send other way

DigitAll

its only on discord, connect patreon with discord, or i can send it to you

Unbound Logic

where is the link to download comfy ui studio on patreon?

DigitAll

Write me on discord

Roberto Strempler

Hey Alex I cannot find the update link in discord

Charles Phillips

Is there a new release coming soon? can we get a toned down workflow for older machines? I have only 12GB VRAM (max available on my PC), with 24GB of regular RAM (Also max I can use). I have tried GGUF, Nunchaku, SVDQuad workflows with custom nodes (all of which crash and the trifecta of ChatGPT, Grok, nor Deepseek have been able to make me waste dozens of hours in total trying "fixes" to get it to work), nothing helps. I even tried going down to 256x256 and it still takes forever or doesn't finish

DigitAll

Just renamed it from 278. I deleted unused workflows

m ab

v256 ? is that a typo

DigitAll

did you connect patreon account with discord account?

Lukas Neumann

I dont find ComfyUI Studio in Discord - when i click the link in your profile it leads me to nirvana

Chris McKellen

I dint see u in my discord at all and really need to update the studio its outdated and won't update the nodes

DigitAll

Connected Patreon with discord?

Chris McKellen

i cant find you in my discord

jonathan sanchez

are you going to install tryton on any version of comfystudio?

DigitAll

ask chat gpt, it will guide you how to install curl

michel sneyers

while downloading ALLn MODELS i get this: Downloading sam_vit_b_01ec64.pth to D:\5-MODELS\0-ALL-MODELS-SCRIPTS\..\..\App\ComfyUI\models\sams... curl: cannot open 'D:\5-MODELS\0-ALL-MODELS-SCRIPTS\..\..\App\ComfyUI\models\sams\sam_vit_b_01ec64.pth' curl: (23) Failed writing received data to disk/application

DigitAll

256 is latest

Trader McBook

Thank you for your excellent work. I can't find ComfyUI Studio v258 link, only Studio v256 link. Where can I find v258 link?

DigitAll

This is the latest

M. Ray White

Is this Revision 258 OR Revision 276?

DigitAll

not supported

Austin Wilson

How well does this run on runcomfy or think diffusion?

DigitAll

Did you connect Patreon with discord? In Patreon settings

Desmond Lee

I can't see all the channels in Discord. Where do I download the update?

DigitAll

Many Patreon accounts are shared in public, my too. Sad story...

LVNDSCAPE

Why discord only now ? I am not a fan of Discord. Too much hazzle with login alltime using this and that account and IP validation.

DigitAll

C:\Studio\Studio v276\App> not good C:\Studio\App> good

Rayeta76

hello again with the v276 version I get a pyhton error the path where I have comfyUI v276 is this C:\Studio Studio v276 but when I start it appears this error in the console ComfyUI Studio v276 ---------------------- Connect your Patreon account to your Discord account to get early access and daily updates. ---------------------- YouTube: https://www.youtube.com/@ComfyUIstudio -------------------------------------------------------------------------------- Updates are available here: https://www.patreon.com/posts/comfyui-studio-111135526 -------------------------------------------------------------------------------- For assistance, feel free to contact me on Discord: my username is "comfyuistudio" -------------------------------------------------------------------------------- C:\Studio\Studio v276\App\python_embeded\python.exe: can't open file 'C:\\Studio\\Studio v276\\App\\ComfyUI\\main.py': [Errno 2] No such file or directory Presione una tecla para continuar . . .

DigitAll

Yes.

Rayeta76

we have a downloadable pdf with the changes

DigitAll

Yes. I renamed it to Depthmap 4

Bill P

did you remove the work flow that created coin design's BAS ? or is it renamed i can not find it now, thanks

Supergonzo2011

Thank you so very much for the hard work. It's a real service you're doing for so many of us. thank you!!

DigitAll

maybe the input image has more objects. or the size is not good

Wills_VFX

PRODUCT BG V2 workflow not working for me : ImageComposite+ too many indices for tensor of dimension 3

Gregory Contreras

Nevermind I got the readme file

Gregory Contreras

Which workflow you recommend me for Background Removal?

DigitAll

use pinochio app to install fluxgym

Мария Павлова

Hello. When uploading photos to FluxGym from the 6-APPS folder, an error appears - "Unable to create tensor, you should probably activate padding with 'padding=True' to have batched tensors with the same length". How can I fix it?

Alexey Trusov

Thank you very much for the new update!

DigitAll

Thanks. Enjoy

Jose M. Aranda

Thank you very much for the update and for the improvements in Hidream and Ltx 0.9.7.

DigitAll

build is comlex to use it on other device

Владимир Кунаков

Dear author, do you have a ready-made file to use on rented gpu's? Your build for windows would be great if there was a Linux version.

Art Szabo

Yes! Also maybe an ongoing series only for subscribers on what can be possible when combining workflows etc. Just ideas...

marwa velan

Would amazing if you did a video on every workflow. Like talking about it, and doing a demo with a brief explanation

Nik

yeees please workflow with upscale video <3

Jose M. Aranda

New version FP8 LTXV 13B 0.9.7 would be great if you could implement it for the next update. It looks like a big change in quality. There are workflows with upscale video on the github page. https://github.com/Lightricks/ComfyUI-LTXVideo

m ab

j'ai une erreur avec une rtx 5080 et lous les workflow standards : CLIPTextEncode CUDA error: no kernel image is available for execution on the device CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. For debugging consider passing CUDA_LAUNCH_BLOCKING=1 Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions

Mohammed Allootah

is there a training for all the workflow . video or pdf

DigitAll

H: 1New folder \ rename this folder, must not contain spacees

Mohammed Allootah

It give me error when downloading. "" Running script: 4X foolhardy remacri [63.92MB].bat Downloading 4x_ foolhardy Remacri.pth to H: 1New folder \v26815-MODELSIO-ALL-MODELS-SCRIPTS....App ComfyUI models lupscale_ models... ** Resuming transfer from byte position 67025055 ""

Kerem Copcu

I try to make copy and try to download but links not working . Is it possible to share another location for installation files ? It's very hard to download ..

DigitAll

No need.

O.M.3

I just format my pc, installed fresh windows 11, do i need install (get and python) to run comfyui studio?

DigitAll

yes. the models are now moved in 2other folders. 0-ALL-MODELS-SCRIPTS MODELS BY WORKFLOW

red lateral

this script's are'n in the package "5-MODELS/Manual download" (here are the main model downloading scripts) 2️⃣ "5-MODELS/Manual download/big models" (

LVNDSCAPE

Regarding your NF4 Flux Workflow i do have a request if possible: Can you please add a Multi LORA Loader to your NF4 Flux template to be activated or not? Regarding ComfyUI Studio Workflow development: I would appreciate SD 1.5 , Pony and Illustrious SDXL workflows you think are in your opinion the best option for creativity one day.

DigitAll

Check your models folder its in App\Comfyui\models_backup. Replace the models folder in the ComfyUI folder.

Laurent Cresson

a lot of missing files just after v268 install lot of checkpoints

DigitAll

Ok will do. But there is option to download models by workflow.

Wolfgang D.

Frame Pack delivers interesting results. However, as with WAN, the generation times on my RTX4060Ti, 16GB are far too long (33 minutes for 10 seconds video). It would be great if instead of the 'Download all Models.bat' there was one without the video models. This would save a lot of time during installation or updates.

DigitAll

Write me on discord

Vahan Terterian

in IC LiGHT 2 the LayerUtility: HLFrequencyDetailRestore and LayerUtility: ImageBlendAdvance nodes are not working, they are missed, and it is impossible to load them by missing node manager. how can i fix it?

DigitAll

Yes

Steffen Kirkegaard

Will it be possible to implement Skyreels? https://www.aisharenet.com/en/skyreels-v2/ https://huggingface.co/Skywork/SkyReels-V2-I2V-14B-540P/tree/main

Many Wayne

I have a generally poor output quality

Paulo Chiovatto

voice clone you can put the brazilian portuguese voice

M. Ray White

A few revisions ago, your installer automatically installed to my desktop a "ConfyUI" icon. It was the start.bat file with the ico file attached. Can you make that unique launch from the desktop feature permanent in all of your present and future builds?

DigitAll

ok

Bolek Lolek

Can you please make a video on how to use your workflow “VIDEO HUNYUAN LONG VIDEO” and the others, and maybe a general video on the Wan theme, a video with short examples on how to use the workflows

DigitAll

Send me screenshot on discord

Nicolas Calabrese

hello, every time i change a variable for example the number of seeds i get this error and I must save the changes and restart Error Cannot read properties of undefined (reading 'beforeQueued')

DigitAll

Just close this message. Its not error

denni dondi

In Hidream workflow, This error appears, what am I doing wrong? Link 365 is invalid, both origin 133 and target 141 do not exist Link 364 is invalid, both origin 133 and target 146 do not exist Link 363 is invalid, both origin 133 and target 141 do not exist Link 362 is invalid, both origin 133 and target 146 do not exist Link 360 is invalid, both origin 133 and target 146 do not exist Link 359 is invalid, both origin 133 and target 141 do not exist 365 is def invalid; BOTH origin node 133 is removed and 133 target node is removed. 364 is def invalid; BOTH origin node 133 is removed and 133 target node is removed. 363 is def invalid; BOTH origin node 133 is removed and 133 target node is removed. 362 is def invalid; BOTH origin node 133 is removed and 133 target node is removed. 360 is def invalid; BOTH origin node 133 is removed and 133 target node is removed. 359 is def invalid; BOTH origin node 133 is removed and 133 target node is removed. Deleting link #359. splicing 3 from links Deleting link #360. splicing 3 from links Deleting link #362. splicing 3 from links Deleting link #363. splicing 3 from links Deleting link #364. splicing 3 from links Deleting link #365. splicing 3 from links Made no node link patches, and 6 stale link removals

Sai Xiong

For those struggling with CFG-zero-star and WanFunControlToVideo missing nodes, I managed to fix it but take these instructions with caution since I don't how it affects ComfyUIStudio fully. Luckily, since we're using git, we can revert back to the branch 3.27 which is originally used by ComfyUIStudio. From my understanding, since the two nodes above are part of comfy-core, we need to update our comfyui to version that's higher than 3.27. So I switched the comfyui repo to use the main branch and voila, the nodes exists. I followed these instructions: Since you are currently on a specific tag (v0.3.27), reverting is very straightforward using Git. Here's the safe process: The Safety Net: The Tag v0.3.27 The key is that v0.3.27 is a named reference (a tag) pointing to the exact commit you currently have checked out. As long as that tag exists (which it does, according to your git tag -l output), you can always return to it. How to Revert (If the Update Goes Wrong): Try the Update: Follow the steps from the previous message: cd A:\Comfy_UI_Studio\v223\App\ComfyUI git checkout master git pull origin master Then restart ComfyUI and test if it works and if the nodes are present. If Things Break or Don't Work: If the updated version has problems (errors, crashes, missing features, nodes still not there, etc.), you can easily revert back to the v0.3.27 tag: Open Command Prompt in the ComfyUI directory: cd A:\Comfy_UI_Studio\v223\App\ComfyUI Checkout the tag: git checkout v0.3.27 You should see output confirming you are now in a 'detached HEAD' state again, pointing at v0.3.27. This means the codebase is exactly as it was before you ran the update commands. Restart ComfyUI: Close the server and run Startup.bat again. You should be back to the version you started with.

DigitAll

thanks will check it and fix it.

Alex Qwerty

Hi! V229 Wan video nodes doesn't work

DigitAll

what error?

Alfonsozm

hello. How can I reinstall the V223 version, this one gives me errors in wan video that the other one didn't give me.

DigitAll

Send me message on discord i fix it

WCMT2020

Updating missing nodes doesn't work for the latest v29 WAN workflows,WanVideoStartEndDecode_2frames WanVideoStartEndSampler_2frames WanVideoModelLoader_2frames WanVideoVAELoader_2frames LoadWanVideoClipTextEncoder_2frames WanVideoStartEndImageClipEncode_2frames CFGZeroStar WanFunControlToVideo

Bruno Solano

Diffusers MV Sampler is not fixed on Workflow Character V2

DigitAll

Download the update 16MB. It will fix it

denni dondi

until 3 days ago it didn't give me this error, today reopening comfy gives me this error... Loading aborted due to error reloading workflow data TypeError: graph.nodes is not iterable

DigitAll

Send me message on discord to help you

Wupi

im always getting a "Download failed after 5 attempts." error while running the '1-DOWNLOAD ALL MODELS" batch file

DigitAll

All output images contain the workflow. Just drag in ComfyUI.

VirtualBite

With this update the costom workflows I made and saved have gone up in smoke :-,( This wasn't the case with previous updates. Note to self and others: make a backup before updating!

DigitAll

Can you open command prompt window in python embeded folder? Or need help?

Manuel Leiton

Hi3DGEN don´t work on me.got prompt Setting attention backend to: xformers Setting sparse backend to: spconv spconv not available Failed to set spconv or not installed, fallback to default. Loading local model: ss_dec_conv3d_16l8_fp16 Loading local model: ss_flow_normal_dit_L_16l8_fp16 Loading local model: slat_dec_mesh_swin8_B_64l8m256c_fp16 Error loading TRELLIS model: No module named 'spconv' !!! Exception during processing !!! No module named 'spconv' Traceback (most recent call last): File "E:\V229\App\ComfyUI\execution.py", line 327, in execute output_data, output_ui, has_subgraph = get_output_data(obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb)

DigitAll

fixed

Nik

UpdateV229 do not work "The file is in the owner's recycle bin"

Fernado Landín Urbina

Sorry for this but the download links are working again. Thank you.

Fernado Landín Urbina

I had some problems with the software and I decided to clean it and to charge it again, but now the download link is desable and I don't know how to reinstal it.

DigitAll

Write me on discord

Fernando Giovanini

Hi, having problems with my fresh installation, it wont start. Getting: TypeError: graph.nodes is not iterable

Tyler

You have to right click and "make a copy" and download it from your own google drive

Tyler

You have to right click and "make a copy" and download it from your own google drive

Iván Holguín

I get this error after running the update script: ComfyUI is outdated, so some built-in nodes cannot be used.

Suresh K. Volam

Sorry, you can't view or download this file at this time. Too many users have viewed or downloaded this file recently. Please try accessing the file again later. If the file you are trying to access is particularly large or is shared with many people, it may take up to 24 hours to be able to view or download the file. If you still can't access a file after 24 hours, contact your domain administrator.

RMonster

Are you sure someone isn't leaking the urls? Can't download either. "Sorry, you can't view or download this file at this time. Too many users have viewed or downloaded this file recently. Please try accessing the file again later. If the file that you are trying to access is particularly large or is shared with many people, it may take up to 24 hours to be able to view or download the file. If you still can't access a file after 24 hours, contact your domain administrator."

Rodrigo Mariano

error wan Video, Missing node type CFGZeroStar WanFunControlToVideo

Javi Sanchi

It won't let me download the update. It says there are too many downloads.

Bill Hancock

I've looked all over the interwebs but I can't seem to find clipvision-for-ip-adapter.safetensors anywhere. Is there another name for it?

DigitAll

sorry. python 3.12 dont support all nodes that studio has.

DigitAll

Hi. Resolution must be max 1024px, and ksampler steps minimum 20

DigitAll

Hi. My workflow don't use sage attention. Choose sdpa in the node instead of sage attention

Laurent

hello, The sage attention module is not detected, how can I fix this? Warning: Could not load sageattention: No module named 'sageattention' sageattention package is not installed.

DigitAll

its working, login to google account, it may resolve it

Bill P

link for updates not working

DigitAll

Send me screenshots of the ssd size before and after generation.

Stanislav Stoyanov

I don't know what is going on when I generate images is taking a lot of space on my SSD although I am deleting all the images I have generated. I am running out of space I keep deleting things from my hard drive. I am going to output folder and delete all the images but that doesn't have much impact if any. How do I solve this what is affecting my storage when generating images even when delete them?

Wari777

Hi, is there a known solution for the verticle banding or squares showing up on the output renders? Sorry if you answered this before.

Deluxe Music Selection

i hope the next update will also be for the 50series graphic cards, i had to install the ComfyUI_windows_portable_nightly_pytorch for comfy to work.

bruno davi

I have this problem with update files Sorry, you can't view or download this file right now. Too many people have recently viewed or downloaded this file. Please try accessing it again later. If the file you're trying to access is particularly large or shared with many people, you may need to wait up to 24 hours before you can view or download it. If you still can't access the file after 24 hours, contact your domain administrator.

Cris

ahh ok haha, that must be it :)

Johnathan B

No, I believe that since some workflows have various sub-workflows inside, that make the 200 🤷🏻‍♂️

Cris

The exact same thing is happening to me, were you able to solve it?

Johnathan B

Hi! I only downloaded the last version and ran the update. I only see 63 workflows, did I do anything wrong? Thanks!

DigitAll

cant be done

William Van Leer

Can Comfy UI Studio be used on the cloud server service, Mimic PC? I can load the work flows but all the nodes are missing. I don't know a lot about it, but before I try to learn, please tell me if it can be done properly and with your permission

Wolfgang D.

Will it be possible to use the GGUF models for WAN2.1?

Gamearc

I reinstalled and here's what i've got Checkpoint files will always be loaded safely. Total VRAM 24576 MB, total RAM 32619 MB pytorch version: 2.6.0+cu124 xformers version: 0.0.29.post3 Set vram state to: NORMAL_VRAM Device: cuda:0 NVIDIA GeForce RTX 3090 : cudaMallocAsync Using xformers attention ComfyUI version: 0.3.22 [Prompt Server] web root: E:\STUDIO\v219\App\python_embeded\Lib\site-packages\comfyui_frontend_package\static E:\STUDIO\v219\App\python_embeded\Lib\site-packages\kornia\feature\lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead. @torch.cuda.amp.custom_fwd(cast_inputs=torch.float32) Traceback (most recent call last): File "E:\STUDIO\v219\App\ComfyUI\nodes.py", line 2147, in load_custom_node module_spec.loader.exec_module(module) File "", line 940, in exec_module File "", line 241, in _call_with_frames_removed File "E:\STUDIO\v219\App\ComfyUI\comfy_extras\nodes_audio.py", line 1, in import torchaudio File "E:\STUDIO\v219\App\python_embeded\Lib\site-packages\torchaudio\__init__.py", line 2, in from . import _extension # noqa # usort: skip ^^^^^^^^^^^^^^^^^^^^^^^^ File "E:\STUDIO\v219\App\python_embeded\Lib\site-packages\torchaudio\_extension\__init__.py", line 38, in _load_lib("libtorchaudio") File "E:\STUDIO\v219\App\python_embeded\Lib\site-packages\torchaudio\_extension\utils.py", line 60, in _load_lib torch.ops.load_library(path) File "E:\STUDIO\v219\App\python_embeded\Lib\site-packages\torch\_ops.py", line 1357, in load_library ctypes.CDLL(path) File "ctypes\__init__.py", line 376, in __init__ FileNotFoundError: Could not find module 'E:\STUDIO\v219\App\python_embeded\Lib\site-packages\torchaudio\lib\libtorchaudio.pyd' (or one of its dependencies). Try using the full path with constructor syntax.

John

Thanks, it would be great if you could provide a mega.nz download link too, but the update is appreciated.

Mike Chandler

Hi! Thanks for this nice Work. Could you perhaps also make a template for rampot? So I can download everything at the touch of a button.

4DXP

Thank you for this. It is a great pack. Browsing through the workflows right now.

DigitAll

The models you can use from the old

Andy Beck

Is der a way to update older version instead of download all new?

Banchert VANTHA

218 update sir

DigitAll

Its ok.

Wolfgang D.

I made a new Install using V200, Update V205 and Update V215. When starting the cmd window shows the error ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts. broken-source 0.8.0.dev0 requires setuptools~=75.2.0, but you have setuptools 75.6.0 which is incompatible. inference-cli 0.22.2 requires requests<=2.31.0, but you have requests 2.32.3 which is incompatible. inference-cli 0.22.2 requires rich<=13.5.2, but you have rich 13.9.4 which is incompatible. inference-cli 0.22.2 requires typer==0.9.0, but you have typer 0.15.1 which is incompatible. inference-gpu 0.22.2 requires onnxruntime-gpu~=1.15.1, but you have onnxruntime-gpu 1.20.1 which is incompatible. inference-gpu 0.22.2 requires requests<=2.31.0, but you have requests 2.32.3 which is incompatible. inference-gpu 0.22.2 requires rich<=13.5.2, but you have rich 13.9.4 which is incompatible. inference-gpu 0.22.2 requires setuptools<=72.1.0,>=70.0.0, but you have setuptools 75.6.0 which is incompatible. inference-gpu 0.22.2 requires typer==0.9.0, but you have typer 0.15.1 which is incompatible. mediapipe 0.10.14 requires protobuf<5,>=4.25.3, but you have protobuf 5.28.2 which is incompatible. py3langid 0.3.0 requires numpy>=2.0.0; python_version >= "3.9", but you have numpy 1.26.4 which is incompatible. torchscale 0.3.0 requires fairscale==0.4.0, but you have fairscale 0.4.13 which is incompatible. However so far I encountered no errors using the workflows.

DigitAll

EMAIL instruction sent. You must connect patreon account with discord. Patreon/settings/more

gersongabrielgm

Gorka Soria

I was asking because I've seen interesting workflows on the internet and of course when I open them in Confifui I have to download custom nodes or something like that. And since the post says that if I download custom nodes these workflows break. I'm sorry, I'm new and I don't know how this works. Sorry.

R P

in audio 4 missing languages i dont see spanish only JP,hi,fr,F5

Alfonso Marquez

Try the reserve link, that one worked for me

TacImpulse

Yeah, me neither. Anyone got a backup link for us? Pretty please?

AirborneSapper708 RoughTerrain12B

lol. All of your workflows are on the “internet” brother .., and ina complex wrapper.,, pot calling the kettle black . It was a valid question. Not a “waste” of time…👍

Gurmeet Singh

I am not able to download the Update v215, Google giving quota exceeded error, even your workaround of copy file is not working. Requesting you support

DigitAll

Yes it's not under video 16. Its VIDEO WAN workflow

Steffen Kirkegaard

Wan video is not under video 16, and where is the archive?

Deluxe Music Selection

oh thank you

DigitAll

On discord you can download it

DigitAll

Its in update v215 archive

Deluxe Music Selection

would be nice only to download the workflow :)

JOHN VOLTAIRE GARCIA

I cant find the video wan workflow. Downloaded the complete package and downloaded all models completely.

Desmond Lee

I updated V200 to V215...update seem fine but haven't tested it. What do I do? What issue I might be facing?

DigitAll

No. Must update 205

Desmond Lee

Can I upgrade from V200 to V215 directly and skipping V205 etc.?

Games_Graphics

any chance you can upload to mediafire instead of google? Since google has that "download quota exceeded..." message. Im not able to download the latest update, and it wont let me make a copy on my google drive, I have 13gb free space.

DigitAll

you must send me screenshots on discord or facebook

kamil

When loading the graph, the following node types were not found ApplyTeaCachePatch FluxForwardOverrider try fix does not work

kamil

(IMPORT FAILED) ComfyUI_Patches_ll

kamil

in source file also missing workflows: Morph 7 V2,Skyreels Hunyuan in VIDEO,Replace with lora flux,BAS RELIEF upgrade,AUDIO 4 sonic lipsync

kamil

Workflow pack: v215 Missing Node Types When loading the graph, the following node types were not found WANImageToVideo. but when i open manager and "install missing nodes" there is no results

주현 한

Can you please make a version that works with Nvidia rtx 5090? I'll be waiting.

DigitAll

Try update ComfyUI in manager menu. If not working try extract Studio v211

bruno davi

hi, on the ACE++ 3 node, it gives me this (missing node type "ApplyTeaCachePatch" and "FluxForwardOverrider"), done with manager (uninstall of the old nodes) then installed new ones but it still doesn't work.

Alfonso Marquez

Hello, nice to meet you, I hope you are well. Do you plan to include WAN 2.1?

DigitAll

ok

Many Wayne

Please make a guide for SKYREELS HUNYUAN IMG2VID. I have encountered many difficulties, and I cannot solve them on my own. I think your guidance would help me. Thanks

DigitAll

Don't use workflows from internet if you don't know how ComfyUI works. It will not work. Jus waist of time

Gorka Soria

Can I ask something? If I download a workflow from the internet and it tells me to download nodes or custom nodes, I can't do it, right? And I don't know if I have the Confifui configured correctly because I think it uses more of my RAM than my GPU and it takes a long time to do anything. I have an RTX 4080super. For everything else, the work you've done is very, very good and I will continue to subscribe to Patreon because all this seems impressive!! Even though I don't know how it works, but I will learn!! hehe

DigitAll

update to v205 then v211. but if you have sonic. just update v211

Thinh Pham Duc

Hello, i’m using V201, how can I update to this new V211 ?

Abdullah Elsayed Bakry Tolba

THANKS BRO

DigitAll

You have received an email how to connect Patreon to discord.

Abdullah Elsayed Bakry Tolba

HI is this Discord server working (comfyuistudio) ? Can't find it at all

DigitAll

Screenshot full error. On discord or Facebook send me

Roman Yuferev

Apply Whisper error in audio 3 folder [WinError 2]

DigitAll

yes download links are in the post

Games_Graphics

is there a update link or do we have to download all over?

LVNDSCAPE

You could recommend the people to login to their google account first and then click your links then there should be no limit as it is for "external" accessing.

Gravity

none of the links work on 205

DigitAll

links active now

DigitAll

links active

DigitAll

hi sorry. there was a bug. now links are available

Frédéric

Hello where are the links

DigitAll

Google limit achieved. Try later.

Jose G Garza

Links are not working!

Kevin O'Malley

is it just me or are theer no links???

DigitAll

Yes 6.7gb. i forget to edit it

Serkan

Hi, update writes 318MB. when try to download it shows 6.7Gb.

DigitAll

Maybe something is blocking the scripts.

Thinh Pham Duc

Hello, I see many issue while downloading All Models : HTTP server doesn't seem to support byte ranges. Cannot resume.

Serkan

Hi can you add, SkyReels and hunyuan image2video please?

DigitAll

Send screenshots on discord

Олег

I had to reinstall everything again

Олег

I don't understand anything... at first everything works, then I try other workflows, go back to those that worked, writes that these nodes are missing, reinstall custom nodes, they are not installed, gives errors...

Олег

In what workflow can I remove unnecessary items from an image? Thanks

Manuel Nixon

Mine still loads with the old version Load Menu. How do I get the new v200 load menu to replace the old one?

Олег

workflow PRODUCT GV 2 there is no IP Adapter the product does not change color there

Олег

you can make a LORA for a product, such as a bag, and then use two LORAS and combine a girl and a bag. Thanks

William Van Leer

I've been using the Vid2Vid V3 and the character stays consistent but the clothing morphs. She's holding a pocket book and then she isn't. Her dress is halter top and then it morphs. Is this just a matter of correct prompt. Any suggestions?

DigitAll

Yes.

Gamearc

So if I have the v196, just download the update and that's it right?

Олег

Damn, I've trained my first two loras, I like it.... Thanks!

gersongabrielgm

segui el proceso del PDF e incluso vi el video de https://www.youtube.com/watch?v=ORcs0gappy8, de Tecnolitas, al ver su video me convencio de suscribirme, al seguir probando, extraje los archivos por medio de Winrar, pero aun asi no me funcionaba y no funcionaba porque le daba que ejecutara como administrador, luego sin hacerlo asi y simplemente dar doble click funciono, disculpa las moelstias, agradezco el apoyo.

DigitAll

Extract 9 archive files with WinRAR app.

gersongabrielgm

buen dia, tengo un problema, al ejecutar 1-DOWNLOAD ALL MODELS.bat, no funciona, sale este mensaje: Error: Folder "5-MODELS/Manual-download" not found. Presione una tecla para continuar . . ., trate de tener una ruta corta, pero aun asi no funciona, pueden ayudarme porfavor?, muchas gracias.

Олег

I thought the Flux Lola training would take place both on and in the pictures obtained on the Flux model, but there is SDXL, very bad

Marie

Can you help me about it please for Runpod?

DigitAll

Yes. You can use any lora

William Van Leer

Can we use Lora models made else where? I had luck with Fluxgym initially and made a Lora as per your video. But I can't get it to work even after clearing video ram. Your project is great as I've said before. Getting a way to use/make new Lora's would be ideal.

DigitAll

Delete spaces in folder names. D:\comfy ui

cessare cessare

After all the instalation and try to open comfyUI i have this error File "D:\Comfy UI Studio\App\python_embeded\Lib\site-packages\torch\__init__.py", line 245, in _load_dll_libraries raise err OSError: [WinError 193] %1 is not a valid Win32 application. Error loading "D:\Comfy UI Studio\App\python_embeded\Lib\site-packages\torch\lib\cusolver64_11.dll" or one of its dependencies. Press any key to continue . . .apreciate your help

Ivan Ruiz

Also locally

Ivan Ruiz

I am using it on Run Pod :/

Dazaeth Garduño

To train loras in fluxgym the minimum card size is 12gb? Or is there any other method to train with 8bg cards? Because I have a quadro rtx 4000 card.

DigitAll

Maybe after using it once, vram is full. So restart pc before using it

William Van Leer

Aurimas Gelumbauskas

Any way to add workflows for product photography/ Background Generation, relighting with flux?

DigitAll

Yes. Watch the lora video on my you tube

Dazaeth Garduño

Do you have a workflow for creating digital influencers?

Khoi Dang

When can you update it to work with the 5090?

DigitAll

no

Khoi Dang

Does this work with the Nvidia rtx 5090 / 5080?

DigitAll

To know i must see workflow screenshot

Steffen Kirkegaard

HI I got this error, can you please tell me what it is? Traceback (most recent call last): File "C:\V196\App\ComfyUI\main.py", line 11, in import utils.extra_config ModuleNotFoundError: No module named 'utils.extra_config' Press any key to continue . . .

DigitAll

Hi. No.

Marie

Hello! Is it possible to use ComfyUI Studio on Runpod?

DigitAll

Under the update download link is a youtube link

Jeff Merilo

Is there a video how the update is done? "Workflow Update Instructions: -Create a new folder in studio root folder. -extract the archive files in that new folder. -Run the "update.bat" script. -Run the "Download all models" script. -Install missing custom nodes if missing."

DigitAll

It's fine. Next update will fiy the version number

DigitAll

Send me screenshots on discord

Desmond Lee

Just updated this "UPDATE V196.rar" (560MB), then only realized I have V190, not V196. Hope it's fine.

Euge If

Hi! What could be the problem? I use all of your models and it gives me this error when I use inpaint and lora? "Requested to load Flux ERROR lora diffusion_model.img_in.weight shape '[3072, 384]' is invalid for input of size 196608 loaded partially 9420.495801635741 9420.431640625 103"

DigitAll

hi. yes

Dazaeth Garduño

Hi! Do you have a workflow with Florence 2?

Gamearc

Sent you friend request

Gamearc

How did you resolve it?

DigitAll

Send me workflow screenshot on discord

Gamearc

I deleted the latentsync node inside the latentsyncwrapper then now comfyui is telling me no latent sync node Traceback (most recent call last): File "H:\COMFYUI STUDIO\V196\App\ComfyUI\custom_nodes\ComfyUI-LatentSyncWrapper\nodes.py", line 31, in import_inference_script spec.loader.exec_module(module) File "", line 940, in exec_module File "", line 241, in _call_with_frames_removed File "H:\COMFYUI STUDIO\V196\App\ComfyUI\custom_nodes\ComfyUI-LatentSyncWrapper\scripts\inference.py", line 19, in from latentsync.models.unet import UNet3DConditionModel ModuleNotFoundError: No module named 'latentsync' During handling of the above exception, another exception occurred: Traceback (most recent call last): File "H:\COMFYUI STUDIO\V196\App\ComfyUI\custom_nodes\ComfyUI-LatentSyncWrapper\nodes.py", line 276, in inference inference_module = import_inference_script(inference_script_path) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "H:\COMFYUI STUDIO\V196\App\ComfyUI\custom_nodes\ComfyUI-LatentSyncWrapper\nodes.py", line 34, in import_inference_script raise ImportError(f"Failed to execute module: {str(e)}") ImportError: Failed to execute module: No module named 'latentsync' !!! Exception during processing !!! Failed to execute module: No module named 'latentsync' Traceback (most recent call last): File "H:\COMFYUI STUDIO\V196\App\ComfyUI\custom_nodes\ComfyUI-LatentSyncWrapper\nodes.py", line 31, in import_inference_script spec.loader.exec_module(module) File "", line 940, in exec_module File "", line 241, in _call_with_frames_removed File "H:\COMFYUI STUDIO\V196\App\ComfyUI\custom_nodes\ComfyUI-LatentSyncWrapper\scripts\inference.py", line 19, in from latentsync.models.unet import UNet3DConditionModel ModuleNotFoundError: No module named 'latentsync' During handling of the above exception, another exception occurred: Traceback (most recent call last): File "H:\COMFYUI STUDIO\V196\App\ComfyUI\execution.py", line 327, in execute output_data, output_ui, has_subgraph = get_output_data(obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "H:\COMFYUI STUDIO\V196\App\ComfyUI\execution.py", line 202, in get_output_data return_values = _map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "H:\COMFYUI STUDIO\V196\App\ComfyUI\execution.py", line 174, in _map_node_over_list process_inputs(input_dict, i) File "H:\COMFYUI STUDIO\V196\App\ComfyUI\execution.py", line 163, in process_inputs results.append(getattr(obj, func)(**inputs)) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "H:\COMFYUI STUDIO\V196\App\ComfyUI\custom_nodes\ComfyUI-LatentSyncWrapper\nodes.py", line 276, in inference inference_module = import_inference_script(inference_script_path) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "H:\COMFYUI STUDIO\V196\App\ComfyUI\custom_nodes\ComfyUI-LatentSyncWrapper\nodes.py", line 34, in import_inference_script raise ImportError(f"Failed to execute module: {str(e)}") ImportError: Failed to execute module: No module named 'latentsync'

DigitAll

uninstsll latent sync node, only latent sync wrapper must be installed

Gamearc

Hi i'm getting this error: # ComfyUI Error Report ## Error Details - **Node ID:** 34 - **Node Type:** D_LatentSyncNode - **Exception Type:** RuntimeError - **Exception Message:** File not found: H:\COMFYUI STUDIO\V196\App\ComfyUI\output\latentsync_rqafi_out.mp4 ## Stack Trace ``` File "H:\COMFYUI STUDIO\V196\App\ComfyUI\execution.py", line 327, in execute output_data, output_ui, has_subgraph = get_output_data(obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "H:\COMFYUI STUDIO\V196\App\ComfyUI\execution.py", line 202, in get_output_data return_values = _map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "H:\COMFYUI STUDIO\V196\App\ComfyUI\execution.py", line 174, in _map_node_over_list process_inputs(input_dict, i) File "H:\COMFYUI STUDIO\V196\App\ComfyUI\execution.py", line 163, in process_inputs results.append(getattr(obj, func)(**inputs)) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "H:\COMFYUI STUDIO\V196\App\ComfyUI\custom_nodes\ComfyUI-LatentSyncWrapper\nodes.py", line 297, in inference processed_frames = io.read_video(output_video_path, pts_unit='sec')[0] # [T, H, W, C] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "H:\COMFYUI STUDIO\V196\App\python_embeded\Lib\site-packages\torchvision\io\video.py", line 271, in read_video raise RuntimeError(f"File not found: {filename}") ``` ## System Information - **ComfyUI Version:** 0.3.12 - **Arguments:** H:\COMFYUI STUDIO\V196\App\ComfyUI\main.py --windows-standalone-build - **OS:** nt - **Python Version:** 3.11.8 (tags/v3.11.8:db85d51, Feb 6 2024, 22:03:32) [MSC v.1937 64 bit (AMD64)] - **Embedded Python:** true - **PyTorch Version:** 2.3.1+cu121 ## Devices - **Name:** cuda:0 NVIDIA GeForce RTX 3090 : cudaMallocAsync - **Type:** cuda - **VRAM Total:** 25769279488 - **VRAM Free:** 22096180210 - **Torch VRAM Total:** 5268045824 - **Torch VRAM Free:** 3018388466 ## Logs ``` 2025-02-06T21:01:34.364431 - 2025-02-06T21:01:34.605431 - Doing inference...: 48%|█████████████████████████████▋ | 11/23 [01:06<00:28, 2.37s/it]2025-02-06T21:01:34.727431 - 2025-02-06T21:01:34.727431 - Sample frames: 16: 0%| | 0/20 [00:00

John McGarrity

Hi, on looking at both download links one is named v196a and the other is v196, the v196a is bigger what's the difference?

Jessica Landa

Is it just me, or are the download links for 193 pointing to a folder with 190a in it?

Skubz

They're in there! Thank you. Also thank you for putting auto reconnect because my internet is spotty as heck. I've lost model downloads from other sites after like 4 hours of downloading and ended up giving up

DigitAll

App\comfyui\models\checkpoints

Skubz

K. The models folder Under comfyui/app/whatever or the one under "v190" (where "download all models" is located)?

DigitAll

Check the models in folder

Skubz

Download all models is saying failed after five attempts on all models but then also immediately saying downloaded successfully. Is it still working?

DigitAll

Sorry for that. Will make warning before update. It must delete all old workflows because if i rename a workflow you will have the old and a new one duplicate.

Wolfgang D.

Unfortunately, the update to V193 did not run entirely without problems. I had saved my own or modified workflows in two subdirectories of the workflows. These were all deleted and I now have to re-import them from generated files. I will be sure to back up my data before the next update.

Miriam Kassandropolis

Those downloads are slower as sleeping turtle from Google Drive. No way to speed them up with file from your webpage? Did this Comfy UI support Stablity Matrix you have all models from all AI Software in one Model Folder only?

DigitAll

The 9 parts always extract in a new folder. WinRAR extract it all at once. No need to do it one by one... Important the names of the files must be same. V190.part01.rar V190.part02.rar

Arbaaz Nasser

where should I extract it. Should I make a new folder in the drive where I have my other comfy ui, pinokio etc folders as well and thn extract them 1 by 1 to that folder which I shall name 'Comfyui Studio'? Am I correct?

DigitAll

Yes. Now use WinRAR to extract.

Arbaaz Nasser

Hello I was wondering whether there is a proper installation guide/Instructions as I am kinda lost. I have downloaded all the RAR files from the v90 folder.

Soundlufs

Of course its good - my question is not about the upscale - its about the "segment" ControlNet used in the workflow. Can it be used with flux?

DigitAll

Thanks. The Flux upscale is now very good.

Soundlufs

Thanks for your great work! I wonder if its possible to add another "2000px Latent Upscale" type workflow, but with Flux? I could`nt find any other workflow using the "segment" processing in the Union ControlNet, but if there is already, or easy way to adjust it, please let me know. Thanks in advance!!

Anwar Hussain

Just thanking you here aswell, appreciate your time and effort, finally working

DigitAll

There is cosmos in studio. Video workflow. But the models for it are in manual download \ big models folder

Anwar Hussain

I've sent a friend request

MarcAndreMueller

Hi, are you planning to add support in the future for Cosmos from Nvidia? You need to update comfy in order to make it work. Should I install a seperate version of comfyui for Cosmos? Since updating comfy might break your workflows right?

DigitAll

Discord to chat. My username is comfyuistudio

Anwar Hussain

How do I do that? Is there a faster way to chat?

DigitAll

I can help you make it work with team viewer pc access

DigitAll

Yes. Extract the 9 archive files in a new folder. Replace the"models" folder from old to new. Use the new one

Anwar Hussain

I did that, same thing, I only have latentsync wrapper installed, I hate ComfyUI honestly it's always a pain, anyway any other solution ? I paid just for the audio, the only thing I wanted is not working really frustrating. Please help me

bruno davi

Hi, I messed up I updated everything instead of just the nodes, I read that you explained to a user to replace only certain files, I can't find the explanation to replace the files without starting from scratch with the installation. Can you help me?

DigitAll

Uninstall latent sync. Only wrapper must be installed

Anwar Hussain

hey ive been scratching my head for the past 2 days, on the audio 3 workflow im getting this error File not found: F:\2025 comfyui studio\V190\App\ComfyUI\output\latentsync_exadn_out.mp4, [NULL @ 0000019ff503bb80] Unable to find a suitable output format for 'F:\2025' F:\2025: Invalid argument Error during inference: File not found: F:\2025 comfyui studio\V190\App\ComfyUI\output\latentsync_tfckg_out.mp4 Traceback (most recent call last): File "F:\2025 comfyui studio\V190\App\ComfyUI\custom_nodes\ComfyUI-LatentSyncWrapper\nodes.py", line 297, in inference processed_frames = io.read_video(output_video_path, pts_unit='sec')[0] # [T, H, W, C] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\2025 comfyui studio\V190\App\python_embeded\Lib\site-packages\torchvision\io\video.py", line 271, in read_video raise RuntimeError(f"File not found: {filename}") RuntimeError: File not found: F:\2025 comfyui studio\V190\App\ComfyUI\output\latentsync_tfckg_out.mp4 !!! Exception during processing !!! File not found: F:\2025 comfyui studio\V190\App\ComfyUI\output\latentsync_tfckg_out.mp4 Traceback (most recent call last): File "F:\2025 comfyui studio\V190\App\ComfyUI\execution.py", line 327, in execute output_data, output_ui, has_subgraph = get_output_data(obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\2025 comfyui studio\V190\App\ComfyUI\execution.py", line 202, in get_output_data return_values = _map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\2025 comfyui studio\V190\App\ComfyUI\execution.py", line 174, in _map_node_over_list process_inputs(input_dict, i) File "F:\2025 comfyui studio\V190\App\ComfyUI\execution.py", line 163, in process_inputs results.append(getattr(obj, func)(**inputs)) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\2025 comfyui studio\V190\App\ComfyUI\custom_nodes\ComfyUI-LatentSyncWrapper\nodes.py", line 297, in inference processed_frames = io.read_video(output_video_path, pts_unit='sec')[0] # [T, H, W, C] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "F:\2025 comfyui studio\V190\App\python_embeded\Lib\site-packages\torchvision\io\video.py", line 271, in read_video raise RuntimeError(f"File not found: {filename}") RuntimeError: File not found: F:\2025 comfyui studio\V190\App\ComfyUI\output\latentsync_tfckg_out.mp4

Alex Qwerty

Unable to join discord,

DigitAll

Try install fluxgym with pinokio app

Damien Watkins

@comfyUI Studio ? Need help ASAP

DigitAll

send me screenshot in private or facebook i need to see

Art Cave

Unable to join discord, Please let me know, I am facing a "Missing Node Problem F5TTSAudioInputs". I tried uninstallation and Installation via comfy manager

Diego Cavali

thx for the audio generator!, been waiting for a long time :), another suggestion would be video maker from a series of images (slideshow) which you can either get the images generated batch uploaded.

Damien Watkins

Hello - I have ComfyUi running beautifully. Having a problem launching FluxGym. It all installed fine but the "RUN" .bat opens breifly then closes - therefore failing to launch FluxGym. Can I edit the .bat somehow to get it to trigger app ? Current prompt is "@echo off cd fluxgym call venv\scripts\activate.bat python app.py call venv\scripts\deactivate.bat cd.. "

Alexandre Massart

I found it on hugging face.

Alexandre Massart

Hi, I can't make cosmo img2vid work, the VAE node is red, looks like I'm missing the cosmo_cv8x8x8_1.0.safetensor, where to get it ?

Nemesis_Film

T2V tutorial. Any image to video workflow?

DigitAll

Uninstall and reinstall

Michael Blendow

Please, how can i install F5-TTS?? FFmpeg is installed.

Nemesis_Film

Need more video tutorials, its so slow with a 4090 RTX for your workflows.

DigitAll

V184 not active anymore. V190 is active

RichardRhal

I can't download the Workflow pack: v184 says I don't have permission?

Bill P

i just downloaded the new v190 and workflows i do not see the install files anywhere like was shown in the video on you tube that got me here also none of the menus show up like on the right side to select any work flows any help??

DigitAll

I can help on discord or Facebook

Filipe Santos

For Video, I also get Missing Node Types When loading the graph, the following node types were not found DownloadAndLoadPyramidFlowModel

DigitAll

screensot please on discord

Filipe Santos

I get this on Audio 3 Installation Error: ComfyUI F5-TTS install failed: Failed to clone repo: https://github.com/niknah/ComfyUI-F5-TTS

DigitAll

Install latent sync node from manager, then uninstall it

종현 한

After Update. (update.bat > download all models). I got an error on AUDIO3 IMAGE TO VIDEO + LIPSYNC workflow. "raise ImportError(f"Failed to execute module: {str(e)}") ImportError: Failed to execute module: No module named 'decord'"

Nemesis_Film

he literally says use Winrar only..

DigitAll

Yes. There are important files.

Kantemir Maykop

I tried it, it didn't work.

Chris Carlson

But i still need to download the 30GB pack first correct?

Tyler

I couldn't either but i found that if i open 7-Zip and then navigate to the files through that software i can open up the first zip file and all the files will appear. Then I just drag and drop them into a folder on my computer

DigitAll

9 parts in sum is 34gb

DigitAll

You can download manually only those models you need. there is manual download folder in 5-Models folder. Run only those scripts you need

Chris Carlson

I wish you could link to individual work flows....i do not need 120 GBs of download when I am only going to use a handful of workflows.

DigitAll

There is a 6-Apps folder. Iin that folder is a fresh ffmpeg installer. Try it

Rafael Luna

I fixed it, the system path was wrong, the dest path should have \bin, like this C:\ffmpeg\bin\ `setx PATH "%PATH%;%DEST_PATH%\bin" /m`

Rafael Luna

I fixed it, the system path was wrong, the dest path should have \bin, like this C:\ffmpeg\bin\. `setx PATH "%PATH%;%DEST_PATH%\bin" /m`

Rafael Luna

I'm having the same issue (error) and I installed it as you mentioned too.

Rafael Luna

I'm having the same issue (error) and I installed it as you mentioned too.

Nemesis_Film

I didn’t change names just extract since it’s like 8-9 volumes

Tobe2d

"UPDATE V190.rar" (560MB) Sorry, you can't view or download this file at this time. Maybe you can add more links ?

DigitAll

ok will do a detailed video

Relatos para escuchar

Ok. So that I understand it finally. I only keep of the old installation the "models" folder under the "App/Comfyui" folder, not the 5-Models nor any other folder. Is that right? I guess the "output" and "input" folders should be kept as well, isn't it? I think you could do a much clearer video with instructions, because in the video I saw this was not clear at all. And therefore, I have been doing it wrong all the time.

DigitAll

If you extract new studio, delete the old one. Never overwrite. Only 1 folder move to the new extraction, the "models" folder.

M. Ray White

I note the size of 190 at "Download the 34GB ComfyUI Studio package" ~ but ~ I am not finding an rar file of this size. Am I missing something here?

Relatos para escuchar

Ok, I see now. But in any case, we have to move all the new extracted folders to the original studio installation. So, I guess that I'm doing it but in the inverse direction, and I just copy the new files with scripts and workflows over the old ones, instead of the old ones over the new ones, prior to downloading the models. It makes more sense to me.

DigitAll

The studio has a "models" folder. So if you extract a new studio 34gb, to not download the models again from internet, you can move the downloaded ones from old studio, to the extracted new studio that jas no models downloaded.

Relatos para escuchar

I still think the wording is a bit confusing. Should we move the new models of the new studio and replace the old models with the new ones recently downloaded? In that case, it should be "move the models from the new studio folder to replace the old ones, isn't it?"

DigitAll

Yes

Richard Mathews

Hi, Ive just updated the v184 last week, can i just use the : Workflow pack: v190 to update?

DigitAll

thanks fixed "Move the models from the old studio to the new studio."

Relatos para escuchar

When you say: "To fix errors: Extract a new studio, replace the "models" from old studio to the new one.", I think it should be the opposite, isn't it? You should replace the old models with the new ones, right?

DigitAll

there is a simple workflow update 500MB. always use that update. but any time you can download the full pack too.

John McGarrity

Thanks for the update. What has changed since v185, and also is there an individual update only file or do I need to download v185 again for the included updates. Again thanks.

DigitAll

It allows but not recommend

DigitAll

The name must be original. Check file names

Nemesis_Film

Nemesis_Film

DigitAll

Maybe try a separate ComfyUI to run this 3d stuff. A clean one.

Kantemir Maykop

Excellent! I faced a problem with generation of 3d models. I tried to install nodes myself, but failed. Is it possible to integrate Comfy3D_Pre_Builds into ComfyUI Studio?

Daniela Vergara

I already installed it, but it must be in a specific location or just with ffmpeg install run as administrator.bat? because it still doesn't work

DigitAll

Download the v185 workflow update again. I added languagees.

DigitAll

In the update post is attached a ffmpeg installer. Install ffmpeg

Daniela Vergara

Hi, I get this error in the audio stream: Apply Whisper [WinError 2] The system cannot find the file specified

JOHN VOLTAIRE GARCIA

https://www.dropbox.com/scl/fi/jwmpxdinkyrc1pqciyhuq/473431807_2643308585861666_973679966026866827_n.png?rlkey=bay7gw3pmt0te4y89vglc04s1&st=pamy27ym&dl=0

DigitAll

Send me screenshot of the ffmpeg\bin folder

JOHN VOLTAIRE GARCIA

yup, I installed it and still have the same error.

John McGarrity

sent on discord thanks

DigitAll

Find me on discord or Facebook to send me private message

John McGarrity

Not sure how to send screenshot to you, but its the default workflow as loaded nothing changed (maybe I need to place some input not sure) error says - got prompt Failed to validate prompt for output 24: * VAEDecodeTiled 51:4: - Failed to convert an input value to a INT value: overlap, , invalid literal for int() with base 10: '' Output will be ignored [INFO] LoadPromptsFromFile: file found 'MORTAL COMBAT CHARACTERS 4.txt' Prompt executed in 0.04 seconds

DigitAll

I attached a ffmpeg installer to the post. Install it. It will fix it

JOHN VOLTAIRE GARCIA

HI! I'm getting this error using AUDIO workflow " Apply Whisper"

aprisma

Ok so update don't work. Understand

DigitAll

teacache works good in the v185 full pack 34gb comfyui. try it.

Marian Cepeha

Thanks a lot! Problem fixed!

aprisma

Great work like always but the new TEACache node creates that and stop generation: Failed to validate prompt for output 325: * TeaCacheForImgGen 602: - Value not in list: model_type: 'False' not in ['flux'] - Failed to convert an input value to a FLOAT value: rel_l1_thresh, flux, could not convert string to float: 'flux' Output will be ignored

DigitAll

delete in controlnet folder the diffusion pytorch model, then go to 5-models/manual download folder, and run the instant id controlnet script

Marian Cepeha

Hi there. When trying to use character v3 workflow I get this error in Ksampler node: This Controlnet needs a VAE but none was provided, please use a ControlNetApply node with a VAE input and connect it. Can anyone help? thanks!

Matt Morton

There is one with 3 different methods in v185.

Manish Jain

The one that has 9 rar files . I am downloading it now . Thanks

DigitAll

Added link 2

Manish Jain

I can't download the file. keep getting file has been downloaded too many times. can you please create one more link

Wari777

HI, thank you for this update! Unfortunately, I keep getting the error that downloads have exceeded its quota for these files, I am also unable to make a copy to my drive.

DigitAll

Hi. Send me screenshot of the workflow and error

John McGarrity

Thanks for this, can you help me - On the 2000px BATCH prompt i get this error - got prompt Failed to validate prompt for output 24: * VAEDecodeTiled 51:4: - Failed to convert an input value to a INT value: overlap, , invalid literal for int() with base 10: '' Output will be ignored [INFO] LoadPromptsFromFile: file found 'MORTAL COMBAT CHARACTERS 4.txt' Prompt executed in 0.03 seconds

Deluxe Music Selection

Amazing.... Thank you for the Watermark Workflow

Lazlo

Maybe not. I just tried it and it does not seem to remove scratches.

Lazlo

Isn't that the purpose of the RESTORE IMG workflow?

Đặng Văn Cường

I hope there will be a "WF" to change the strong backgroup, you change the outfit as requested

Ezequiel Casas

Excellent update! Is there a workflow, or maybe it can be implemented in the future, to repair the scratches and tears of an old image? it would be the ideal complement to the image restoration workflow! Perhaps also an add-on to correct the desaturated or yellowish colors of color photos from the 70s to the 90s. Thank you!

DigitAll

Thanks. Enjoy)

Alex Young

Just joined and have to say love the work! everything works perfect on my 4080 Super :)

Deluxe Music Selection

hope there will be a watermark workflow in the near future :)

DigitAll

Detailed install description fixed.

LVNDSCAPE

Thx for your release workflow.

DigitAll

Download the 32gb v171. Then update the workflows v172. Replace the models from your old ComfyUI to the new.

LVNDSCAPE

Hi your latest v172 Info tells Detailed Studio Installation: Download the Studio v168 package in 9 parts to avoid interruptions. --- My latest release is v161 from Dec1 and i did not receive any update e-mail for a v168 Version and i did not see anything here for a v168. And where could one even download this since you refer downloading this? Thx for the v172 Version although the full release tells v171 if i click download. Is there everything correct with the full v172 release being tagged v171 ?

DigitAll

Lora trainer app

Lazlo

What is FluxGym install.zip?

gary

on the options to train with certain models and you select it then it will have to download it only one time and it may appear to be stuck but if you open up the terminal to see. these files on the flux dev are like 20 megabytes which may take a while

DigitAll

Add friend request. Username is comfyuistudio

silfras joeld

I have discord installed, but I can't find you on discord. How do I chat with you on discord?

DigitAll

Write me on discord or Facebook

silfras joeld

Good morning I just signed up for Comfyui Studio, I wanted to know how do I ask questions about installing and using the workflows?

DigitAll

Link to the model download is in the white node. But use gguf. Compact is old

marek kazimierz

INPAINT PRO 10 SIMPLE INPAINT FLUX DEV COMPACT no model in module "Load Checkpoint: flux1CompactCLIPAnd_Flux1DevFp16.safetensors"

Anthony

Hi...I probably should have checked this before joining. Can I use this effectively with MimicPC or does it have to be run locally?

DigitAll

Download it manually. The flux model

ruben sebban

I tried installing fluxgym but it keeps getting stuck at this code: license_items=['license: other', 'license_name: flux-1-dev-non-commercial-license', 'license_link: https://huggingface.co/black-forest-labs/FLUX.1-dev/blob/main/LICENSE.md'] license_str = license: other license_name: flux-1-dev-non-commercial-license license_link: https://huggingface.co/black-forest-labs/FLUX.1-dev/blob/main/LICENSE.md no samples" Do you have any idea how to fix that?

DigitAll

Send me screenshot of root folder. And update folder

종현 한

Hi. Great work. When running Download-All-Models script I got some error Could you please help? // Warning: Failed to open the file C:\Users\user\Downloads\ComfyUI-Studio-v88\UPDATE Warning: V165\files\5-MODELS\Manual-download\..\..\App\ComfyUI\models\controlnet\animatediff\control_v1p_sd15_qrcode_mo Warning: nster.safetensors: No such file or directory 0 689M 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0 curl: (23) client returned ERROR on write of 15554 bytes Downloading control_v1p_sd15_qrcode_monster.safetensors to C:\Users\user\Downloads\ComfyUI-Studio-v88\UPDATE V165\files\5-MODELS\Manual-download\..\..\App\ComfyUI\models\controlnet\animatediff... % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 1165 100 1165 0 0 5470 0 --:--:-- --:--:-- --:--:-- 5521 //

에밀

Now the loading is stuck at "23% - FLUX MODEL" after hitting Queue Prompt and no matter how long I wait, its still at 23%. It worked before but after some time its not working anymore. What can I do?

에밀

it works now, thanks!

DigitAll

hi download the models. The flux dev fp8 model download link is in the white node.

에밀

i get this error when trying to queue promt: * UNETLoader 440: - Value not in list: unet_name: 'flux1-dev-fp8.safetensors' not in ['IC-Light\\iclight_sd15_fc.safetensors', 'hunyuan_video_t2v_720p_bf16.safetensors', 'lotus-depth-g-v2-0.safetensors'] Output will be ignored Prompt executed in 0.03 seconds what can i do?

Silb3erbart 1

ok will try that. Thank you.

DigitAll

Yes. But only the 32gb. The models folder you can replace from old one.

Silb3erbart 1

so... im afraid you suggest DL the latest complete package and re-downloading all 120 GB models new ? -.-... well i could backup the 123 GB modelfolder...

Silb3erbart 1

the older one but updated to 163

DigitAll

Hi. Which version of studio you using. The 32gb v161 or older one?

Silb3erbart 1

hmmm... on both vid2vid workflows i get the message about missing note "ACN_AdvancedControlNetApply_v2". I cant install this missing note as comfyUI cant find it. Manager says: "(IMPORT FAILED) ComfyUI-Advanced-ControlNet" Any help ?

DigitAll

5-MODELS folder has scripts. try those. or use Manual download folder scripts.

aprisma

Is there a way to enter API-Key from Civitai? Or what is the reason that much files fail to load? Running script: Flux-uncensored-lora.bat Downloading uncensored-flux-lora.safetensors to C:\AI\ComfyUI-Studio\5-MODELS\Manual-download\..\..\App\ComfyUI\models\loras\FLUX... % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 1124 100 1124 0 0 7587 0 --:--:-- --:--:-- --:--:-- 7698 100 655M 100 655M 0 0 29.3M 0 0:00:22 0:00:22 --:--:-- 29.8M Download failed after 5 attempts. Script "Flux-uncensored-lora.bat" completed successfully.

DigitAll

Hi. You can experiment with it. There will be no difference in work

alex47me

Is it safe to update ComfyUI, python_embedded, git-portable and nodes to latest versions? Is there any reason why you're using python 3.11 instead of 3.12 or even 3.13? Thanks

João C R Fonseca

At the end of the post after "Need Help?" this: FluxGym install.zip IC LIGHT.json VIDEO VID2VID V2.json PRODUCT BG V2.json VIDEO VID2VID V3.json

DigitAll

Which one?

João C R Fonseca

Hello!! I have a question that may seem silly, but what should I do with the attachments? I'm going to download them and then, where should I save them?

aprisma

Question to Update please: Download and extract "UPDATE V163.rar" -- OK into the "ComfyUI Studio V***/UPDATE V163/" folder. That means I make a folder "update V163" in my comfyui-Studio-Main-folder, right? Run the "update.bat" script from the extracted folder. - OK Run the "Download all models" script (the one in my update V163 folder) This will do a full installation of the models but NOW TO Update V163 Folder. It also not download the delta - it downloads all again. So at tjhe End I have in Update v163: 1-WORKFLOWS, 5-MODELS, App, 1-DOWNLOAD ALL MODELS.bat, Start.bat. .... Should I then copy the content in files-folder manually to the main directory?

Ron Harris-White II

Getting error about Florence2ModelLoader not loading, and LLM folder not being created.

Ron Harris-White II

I can install it, I was curious if there was a way to avoid multiple installs of ComfyUI.

DigitAll

No way to use other platforms with studio workflows. I can help you install if needed. Write me on discord or Facebook i will guide you

Ron Harris-White II

Yes, I am

DigitAll

Hi. You using windows?

Ron Harris-White II

I am using Stability Matrix for ComfyUI. Is there a simpler way to inxtall your workflows into ComfyUI using that platfom?

LVNDSCAPE

Ok fine. Beside this exprience i do not receive any errors in the console starting the Studio which is great. The LTX workflow works great thx a lot :)

DigitAll

Delete the vae-hunyuan.pt script. I fix it in next update.

DigitAll

Write me on discord or Facebook i will help

LVNDSCAPE

Yes the model is there but the 1-DOWNLOAD ALL MODELS.bat always stops with that. It is same in 159. I deleted the vae-hunyan PT, had another 1-DOWNLOAD ALL MODELS.bat run and it is again closing the CMD command at this position same as it is in 161. Is there more to run than "vae-hunyan.bat" ? That is the question and if there is a issue with the model downloader script. There is no completed info to show one that all is fine.

DigitAll

Check ot in models\vae folder. Maybe it's there

LVNDSCAPE

I) do rceive same errors Running script: vae-hunyuan.bat Downloading vae-hunyuan.pt to \ComfyUI_Studio_v161\5-MODELS\Manual-download\..\..\App\ComfyUI\models\vae... ** Resuming transfer from byte position 986000558 % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 1120 100 1120 0 0 6244 0 --:--:-- --:--:-- --:--:-- 6327 100 49 100 49 0 0 199 0 --:--:-- --:--:-- --:--:-- 199 Download failed after 5 attempts. Script "vae-hunyuan.bat" completed successfully. It always stops at the same position. I used the full 1.61 install where i moved my model folder over from v159.

DigitAll

Hi. Your internet connection is stable and fast?

MoGrafix

Hi, when I run the "Download All Models", I keep getting "Download failed after 5 attempts" errors.

DigitAll

try using the new v161 32GB studio

DigitAll

Hi. Thanks. Top is only poses. You can use checkpoints but you must know the settings in Ksampler for the new checkpoints.

aprisma

Wow fast :-) I saw that i can use the extra_path file without problem. Thank you. I have all internal NVME SSDs. Great workflows. Thanx

DigitAll

Hi. Thanks. Try it. Just use internal hard drive only.

aprisma

Hello. Great work. Just a quick question. Does the design of ComfyUI-Studio supports also differnt paths? I have all my additional models on another SSD. Thank you. Have a great day.

Andre Silva

Hello, great work! On Character v3 workflow, the top row of images, how would you use those? after i create my dataset, is there anything else i need to do with the top row? Also if i want to install a different chkpnt, do i just manually copy it to the standard models folder in comfyui or do i need to place it in a different location to use it in the Character v3 flow?

DigitAll

Show me screenshots

Work Official

I fixed it. But now I'm getting the InstantIDFaceAnalysis error

Claude GIROLET

i've got this error when i tried to use character v2 to obtain picture to use flux gym to train a lora DWPreprocessor 'NoneType' object has no attribute 'get_providers' i'm using v159 (updated from v157) and i've re downloaded all the models, anyone knows what it means ? i'm using an geforce 3060 wth 12 gos of vram

DigitAll

It's not possible. Wich extractor? WinRAR?

Work Official

The extractor is asking for a password

DigitAll

Write me directly in private

Bold

I HAVE THE SAME PROBLEM - V159

DigitAll

Must see screenshots. Send me on Facebook or discord

Daniela Vergara

Hello, I run the FluxGym .bat files in order but the application does not open.

O.M.3

ok. thank you i will try it

DigitAll

Allocation on device. It's meaning the VRAM is not enough. Try lower resolution and length.

O.M.3

https://ibb.co/4Sb3fqS

O.M.3

https://ibb.co/RDDhv3g

DigitAll

Send me screenshot of the models\vae folder.

O.M.3

yes, i tried 2 times

DigitAll

It's not working?

O.M.3

thank you for the aweasom workflow, i'v tried hunyuan, but after the generate finish it shows "VAEDecodeTiled" and in the top menu it shows red line and writtn " torch.cudaOutOfMemory Error 73 VAEDecodeTiled"

Shai Bitton

THANKS IT WORKED

DigitAll

hi. you internet connection is not stable, th emodel Realvisxlv4 was not downloaded fully. Go to "5-MODELS/Manual-download" folder. run the script for "realvisxlV40_v40Lightning.bat". Delete the other model that is not downloaded fully.

Shai Bitton

hi all, When im using Character V3 workflow, im trying to run prompt after uploading an image. then i get an error message "CheckpointLoaderSimple Error While desarializing header: MetadataincompleteBuffer"

LVNDSCAPE

This works good thx. With my 3080 i have to lower the batch size to 8 in 1080x1080 and this works just so good. The ComfyUI Studio install process has been a very nice experience. Amazing.

DigitAll

its because 1024x1024 px image size, try change to 1026 or 1080

LVNDSCAPE

Character v3 workflow There are Getty Images Logos / Credits Text in all generated pictures. The Getty logo is always same position middle and there is some other additional mini text / logo is on left bottom corner.

LVNDSCAPE

Hi thx a lot. Downloading the v159 now :)

DigitAll

Link 2 added.

David Fischer

It isn't letting me download or "make a copy". I have 150 can I update to 159?

DigitAll

try right click on it and choose 'make a copy to your google drive'. but now i make a new link, and let you know if ready.

LVNDSCAPE

Hi i cannot download the files since your Google Drive tells the download limit has been exceeded.

DigitAll

Yes. Yes. Yes

Laga Mahesa

Doing a fresh install 154. 1) can I do update 158 without doing 157 first? I don't see a link for 157 that's mentioned above. 2) can I copy the models from my existing 142 folder? 3) could I install 158 in my 142 folder instead?

Manuel Fernandez

Thank you for the quick reply, I tried, I ran the updated script for V158.rar, is there anything else I need to do? Or something I am missing? I appreciate your help!

DigitAll

Download the model with the scripts.

Manuel Fernandez

Hi I get the following error when running the ID+POSE FACE DATASET FOR FLUXGYM workflow: Failed to validate prompt for output 8: * InstantIDModelLoader 22: - Value not in list: instantid_file: 'ip-adapter.bin' not in []

DigitAll

Will make. If you need help write me on discord

Marian Cepeha

Any chance for a short tutorial on how to use Character V2 workflow?

Marian Cepeha

Thanks. I did but it didn't work. What worked for me: created App\ComfyUI\models\llm_gguf folder. Manually downloaded Mistral-7B-Instruct-v0.3.Q4_K_M.gguf and put it in llm_gguf folder. Maybe somebody will find this fix useful.

DigitAll

Run download all models.

Marian Cepeha

Hi. I'm trying to use the Character v2 workflow. I'm getting this error. Any idea how to fix? thanks for all your hard work! Failed to validate prompt for output 478: * Searge_LLM_Node 476: - Value not in list: model: 'Mistral-7B-Instruct-v0.3.Q4_K_M.gguf' not in [] Output will be ignored

Neil Biondich

I downloaded and installed your version of comfyui. When I run it (after updating and all), it displays workflows, but all workflow nodes are not connected and won't connect to each other. No amount of dragging connect points work in any workflow.

DigitAll

Write me on discord or Facebook i help you install it

DigitAll

Yes. Use 11.8

Frederic Chauveau

Ok... I have the 12.4: is it the problem ?

Charles Phillips

I have had nothing but trouble after updating (AND fresh installing). Everything from multiple python errors, getting Comfyui to run, and the following issue...Missing Node Types When loading the graph, the following node types were not found Text Concatenate UnetLoaderGGUF DualCLIPLoaderGGUF

DigitAll

NO NEED cuda toolkit for ComfyUI. If you install CUDA must be the 11.8

Frederic Chauveau

What version of CUDA are you using? I am encountering the following error: [ReActor] 02:06:37 - STATUS - Analyzing Source Image... 2024-12-10 02:06:38.6653699 [E:onnxruntime:Default, provider_bridge_ort.cc:1862 onnxruntime::TryGetProviderInfo_CUDA] D:\a\_work\1\s\onnxruntime\core\session\provider_bridge_ort.cc:1539 onnxruntime::ProviderLibrary::Get [ONNXRuntimeError] : 1 : FAIL : LoadLibrary failed with error 126 "" when trying to load "O:\ComfyUI-studio-v150\App\python_embeded\Lib\site-packages\onnxruntime\capi\onnxruntime_providers_cuda.dll"

DigitAll

Make a back up. And try it

guni

I am currently in V88, can I just use update to V150 without downloading the whole 30G file?

M. Ray White

Brought my mobile workstation up to 150! Thank you for your patience and help!

DigitAll

Inpaint outpainting with flux fill. Redux in gguf and flux original

Miriyala Praveen Kumar

what are the changes on 150v?

DigitAll

Hi. No. There is the update v150 1GB archive. Instructions in this post for update.

Muhammad Hussain

Hi, What if we already have the previous version with v142 update? does v150 requires fresh installation? Sorry, couldn't understand the process.

Madara Uchiha

Oooh I will try soon, thank you

Hussein

hi what the update for this version? and there any update in character sheet workflow?

DigitAll

It will be in the next post

Imad Elorf

Thx for great work, but it will be great if you add a Changelog for every update, so we test the newest features instead of navigating into every workflow and checking whats new on it.

DigitAll

added.

DigitAll

Yes

Madara Uchiha

Hello, are you planning on adding Flux Redux/Advance?

Milsen Ace

Ah ok no problem lucky me have also a windows nvidia machine :)

DigitAll

Hi. It was optimized on windows and for Nvidia vcards. Full potencial only on windows.

Milsen Ace

Maybe this was already answered but can i Use ComfyUI Studio also on a mac?

DigitAll

There is a link in the white node.

Manuel Nixon

Hi, how are you doing? Where might I find the "Flux1CompactCLIPand_flux1devfp16.safetensor" needed for the 'Flux CN Depth workflow'? It hasn't downloaded via the "Download All Models". I'm not able to execute the workflow. Thank you.

guni

how to add the discord? I try to add friend comfyuistudio

DigitAll

Write me on Facebook or discord to check your models. maybe incomplete download of model. Try run the download all models once again. It will resume download models that are not downloaded completely

guni

yes. any style and depth map working properly, only the openpose got an error

DigitAll

with default settings and images?

guni

Hello try Cute You 5 workflow, the cute you openpose woman gives me error: size mismatch for proj_in.weight: copying a param with shape torch.Size([1280, 1280]) from checkpoint, the shape in current model is torch.Size([1280, 1664]) any idea what is happening? thanks

DigitAll

Almost same

Hussein

Thanks ill complete the trip of downloads and hopefully everything work well but I note the scripts in version 88 the same in in last update ?

DigitAll

open scripts with text editor, there is link, name, and place to save it.

Hussein

Hi I started download scripts manually cuz I’m already have some of them on my other Comfyui, but in most cases the CMD not complete downloading and when try open it again not working cuz of that I downloading them from Hugging face one by one and added in your models directory but the files have different name I need rename them with the same script name or this not different?

guni

thanks for the info

guni

ok it kind of works... may be can try to figure out how to change the timing of the morphing as currently seems abit too fast

DigitAll

My ComfyUI is portable. So you can have as many you want. It uses own python library

DigitAll

Yes. Prompting must be related to the picture

guni

In worflow Morph 2, is there a way to kind of controling a more consistent face? I put a reference photo of female asian woman, but it's morphing to western female :D

guni

Thank you for the update. I have my existing comfyUI in D drive (not portable version) and want to install comfyUI studio in E drive, will this affects the exsiting comfyUI in D drive? should I Just run the bat file of ComfyUI studio? do I need to run virtual environment with ComfyUI studio?

John Elder

Great! Thank you

DigitAll

Today will make it.

John Elder

Do you gave a server on Discord?

RudyVRGuy

That fixed it thank you

Hussein

my question now im already install Comfyui what add models from comfy UI workflow or install it in different drive after download it? and if there any pdf or video explain installing steps ill be appreciate that :)

DigitAll

run the Download all models script to download the models. if it not fix it. write me direct i will help

RudyVRGuy

I have the same issue when ever i use any workflow that uses the Load Lotus Model node. - Value not in list: model: 'lotus-depth-g-v2-0.safetensors' not in ['lotus-depth-d-v-1-1-fp16.safetensors', 'lotus-depth-g-v1-0.safetensors', 'lotus-normal-d-v1-0.safetensors']

Imad Elorf

Im preparing a txt file reviewing all workflow's and ill send it to you in discord when done.

Imad Elorf

In "Bas relief 5" Workflow and many others, besides there is a newer version of lotus-normal-d-v1-0.safetensors in the net you can update.

Albert Bardewyck

Once again great cinema Alex!

DigitAll

Which workflow?

Imad Elorf

- Value not in list: model: 'lotus-depth-g-v1-0.safetensors' not in list, if you can add it on auto download .bat :)

grumpy grumpy

Ohhh brother, what an amazing update!! Absolutely brilliant, the features you have added and the updates are first class. Your method of installing updates and working with Comfyui are the best I have seen. So much so I have unsubscribe to all other patreons that deal with Comfyui. Nobody come even close. Super happy to be onboard this OceanLiner

Jeramy Udy

Glad I just got a 4TB hard drive. I have been switching between Pinokio, Stability matrix, and Comfyui studio, and seem to have a whole lot less trouble running your program than the other ones. You are awesome.

Nobodys Hero

142 wow ty

DigitAll

yes. download the update file v142.

Nobodys Hero

Is v96 released ?

justin ruscheinski

perplexity ai works wonders too

grumpy grumpy

Can't wait to see the updates.. all of your workflows are fantastic!

Nobodys Hero

I have to say I am very impressed with your updates. Do you have a omnigen workflow? pyramid flux video looks good hehe AI is ever evolving, good stuff keep up the fantastic work :) Look forward to see V96 tommorow!

DigitAll

Tomorrow will release the v96. Lots of new workflows

Nobodys Hero

whats new in v88

DigitAll

Yes

Richard Mathews

HI, can I update from an older version by installing and then copying the models folder from the old install and then downloading new models in the new install?

Dhizzylab Gon

Hello. Problem solved. It was coming from my winrar, old version. I got a new one. Thank you for your reactivity.

DigitAll

maybe you have renamed some parts, or incomplete download.

DigitAll

Hello. only this can unzip: https://www.win-rar.com/predownload.html?&L=0

Dhizzylab Gon

Hello. I get the error message below when I try to open the files with winrar: H:\COMFYUI-STUDIO-V88.PART1.RAR THE ARCHIVE IS EITHER IN UNKNOWN FORMAT OR DAMAGED I also tried with 7zip, same result.

DigitAll

Ask me first if you need help. )

Soufiane Zahir

Thanks you saved me a lot of headaches 🙌😊

DigitAll

Write me on Facebook or discord to see your files.

wilBr

yes.. I did.... should I do it again?

DigitAll

Hi. Yes. There is an Eye icon on the right. Click it. That's it

Soufiane Zahir

Hello my friend, is there any possibilites in Comfyui settings to show the hidden lines? I tried everything with much succes :( thanks in advance

DigitAll

Hi. You must download all models. With the 1-DOWNLOAD ALL MODELS script

wilBr

Error with "flux dev 8" workflow "flux dev" group: Failed to validate prompt for output 325: * CheckpointLoaderSimple 321: - Value not in list: ckpt_name: 'flux1CompactCLIPAnd_Flux1DevFp16.safetensors' not in ['AbsoluteReality_1.8.1_pruned.safetensors', 'SUPIR-v0F_fp16.safetensors', 'flux1-dev-bnb-nf4.safetensors', 'realvisxlV40_v40LightningBakedvae.safetensors', 'sd3.5_large_fp8_scaled.safetensors'] Output will be ignored Prompt executed in 0.00 seconds

Tyler

Sent you a DM of a screenshot. Its just showing my zip file doesn't even have the main BAT files to launch ComfyUI

DigitAll

Show me the archive files. Screenshot me. In direct message.

Tyler

I made a copy of all parts of the v88 to my google drive and downloaded and fully extracted with no errors besides the 2084 onnx model but i downloaded that. However, I'm missing the 4 main batch files in the zip files. The START, INPUT IMAGES, OUTPUT IMAGES, and DOWNLOAD ALL MODELS batch files in the main folder...

RuggedPineapple

Im requesting a linux version too. This pack was the reason I signed up here only to find out its not compatible is a huge downer. And its not a chatgpt convert. These scripts rely on bundled windows executables throughout.

Mert Durmazer

Is there a way using this workflows in mimicpc comfyui?

Claude GIROLET

hello, what are symlinks and how to use them ? thanks in advance for helping me.

Jeramy Udy

I have been making symlinks for things like the wildcards.

Claude GIROLET

hello could you make the wildcards availabe in any worflows ? (i know this is a lot of work) because there are so many worflows where wildcards would be useful, or at leastg make one video on how we can modifiy worflows to add them by ourselves without breaking your work

DigitAll

hi chat gpt can quick convert them for you.

jc

Any chance you have a version of the bat scripts for shell or bash? I'm pretty sure this is a long shot, but thought I would ask prior to working on getting them converted.

DigitAll

write me on discord or facebook, send me screenshots

Ora Davis

hi i am getting this error: "File "pathlib.py", line 1044, in open FileNotFoundError: [Errno 2] No such file or directory: 'C:\\Users\\orada\\Desktop\\AI\\App\\ComfyUI\\custom_nodes\\ComfyUI-tbox\\..\\..\\models\\annotator\\LiheYoung\\Depth-Anything\\.huggingface\\download\\checkpoints_metric_depth\\depth_anything_metric_depth_indoor.pt.e25a96605a2ea2dbbe57308d024e89a803fa87fe981bd51d4cee245f1de57977.incomplete' " I see the file in the location but it still says it cannot find it. Any help would be greatly appreciated.

DigitAll

Write me on discord or Facebook i wii help you

Hakan Bulut

how to update v63 to v 86?

Thái Xuân Anh

I really like your workflow. I do wedding photos, you help me with the background change process with ipadapter. Sorry I'm Vietnamese, my English is not good

Thorfs

It is working now :) Hat to start in administrator mode, because it wanted to create a folder under users. Strange, but working. Thanks :)

DigitAll

Check the models in this folder: v86\App\ComfyUI\models\vae

Thorfs

I did it. Should I delete the safetensors and download them again?

DigitAll

Download all models. With the script. 1-DOWNLOAD ALL models

Thorfs

Hi, I am testing the img2Vid Workflow and I get an error message. Does anyone have the same issue? Failed to validate prompt for output 53: * VAELoader 85: - Value not in list: vae_name: 'vae-ft-mse-840000-ema-pruned.safetensors' not in ['taesd', 'taesdxl', 'taesd3'] * CheckpointLoaderSimple 564: - Value not in list: ckpt_name: 'AbsoluteReality_1.8.1_pruned.safetensors' not in [] * ControlNetLoaderAdvanced 127: - Value not in list: control_net_name: 'animatediff\control_v1p_sd15_qrcode_monster.safetensors' not in [] * ADE_LoadAnimateDiffModel 87: - Value not in list: model_name: 'AnimateLCM_sd15_t2v.ckpt' not in [] Output will be ignored Prompt executed in 0.05 seconds Thx in advance

DigitAll

? Describe please

Thái Xuân Anh

Hi. Can you please tell me if downloading 46 workflows is enough?

Kusacubari

It was my antivirus, ... turned it off and loads now commencing. Looks superb, cant wait

Kusacubari

Hi, all looks fantast... im struggle to get any models dloading with the .bat files this is a snippet from the cli... curl: (35) schannel: next InitializeSecurityContext failed: CRYPT_E_NO_REVOCATION_CHECK (0x80092012) - The revocation function was unable to check revocation for the certificate. Download completed. anuy ideas to fix, thanks

DigitAll

ok. I attached to this post.

oscar DCDC

hey! i fucked up one of your workflows and i have to downlaod the all package again. Is it possible to have a link only for the workflows?

grumpy grumpy

Pity I cant post the image I just did, took an 8,704 image totally small image upscaled it, converted it to a carving, then to a bas relief soo soo soo brilliant.

grumpy grumpy

absolutely brilliant, love the new update, bas reliefs bonus and works fantastic, I can use those with my laser and cnc equipment genius.. ty ty ty...

DigitAll

It's the gguf 7 workflow. Delete and re download the gguf model. It will download the schnell model and will use in gguf 7 workflow

Tobe2d

Cant find Flux schnell 4 steps up to 1500px you mentioned. what is it called?

DigitAll

No. Not for now.

guni

Thanks got it. It works. I notice the result is actually changing the likeness of face to different people. Is there a setting to minimze that?

DigitAll

In the input folder

guni

Also try to run video img-to-video I have an error message that animation.mp4 failed. where can I get that animation.mp4 ? thanks

guni

14 second

Ora Davis

Fantastic!

DigitAll

how long is the video?

guni

I tried to run the Faceswap video, I dont see the output when its done?

DigitAll

i added a torrent file to this post. Try it.

DigitAll

send me screenshots on discord. Username: comfyuistudio

Peter Uwe Reif

Link 3 doesn't work ! Downloadlimit

fatih uzun

"Warning: Missing Node Types When loading the graph, the following node types were not found: XlabsSampler No selected item Nodes that have failed to load will show as red on the graph." I got this error what should I do?

Peter Uwe Reif

Link 1 and 2 and 3 Download not working !!!! Download Limit.

Chintalapati Vatsal

Thank You! THat worked

DigitAll

not now. ints not constant, but will search for implementation

GhØst•WØlf

Thanks ! It works ! :)

Roman Yuferev

is it possible to add retouch effect to sequence of images or video?

DigitAll

Link 3 available. Test it

DigitAll

Link 3 available. Test it

DigitAll

Link 3 available. Test it

DigitAll

Link 3 available. Test it

DigitAll

Link 3 available. Test it

ricardo rizzutti

I can't download the 5,6,7,8 limit it is pointing to

DigitAll

i'm working on link 3.

DigitAll

i'm working on link 3.

DigitAll

ok. i'm working on link 3.

Chintalapati Vatsal

Both the download links have exceeded their limit. Can you put them onto a dropbox link or some such?

GhØst•WØlf

Download limit problem with the two links. I can't download part 7. :(

Peter Uwe Reif

The same problem, download limit on both links.

Rul Cspm

Succes downloaded via link 2. Tqvm

DigitAll

Try Link 2

Денис Балацкий

you have a download limit from google drive, fix it

DigitAll

Download the 28gb v86. Replace the models folder from v63 to v86. With Ctrl+x and ctrl+v keyboard combination.

Hadi Nooradini

how update v63 to 86 ?

Hadi Nooradini

me2

TheAnim8tor

same

Rul Cspm

hi , i tried to download v86 but the download quota exceeded for this file

Tobe2d

How to update from v63 to v86? can you add update like you did for v52 to v63? it's so slow downloading all these huge files and already have them

DigitAll

Flux schnell 4 steps up to 1500px . And many more.

Tobe2d

Will the next update include Stable Diffusion 3.5 Large? And maybe more flux workflows?

DigitAll

Wait a little. New update will be released today.

red baron

Hey, thanks! Could you also provide the update archive as well to avoid re-installing the whole Studio setup?

ricardo rizzutti

thanks thanks

DigitAll

Hi. Thank you. Uninstall cuda software.

ricardo rizzutti

hello everything is fine!! I really liked the work. Now I have this error and I couldn't fix it "ReActorFaceSwap D:\a\_work\1\s\onnxruntime\python\onnxruntime_pybind_state.cc:636 onnxruntime::python::CreateExecutionProviderInstance CUDA_PATH is set but CUDA wasn't able to be loaded. , and that your GPU is supported."

DigitAll

Hi. It's not in the v63. Download it from the attachments in this post. Put it in the "1-WORKFLOWS" folder. Run the script in this folder "0-Copy workflows to ComfyUI". Press Refresh button. It will be in the list.

Chris Ife

Cute you workflow isn't loading...

DigitAll

Thank you. Never delete models folder.

Haunted Valkyrie

Thank you for the quick response. It was actually the controlnet folder, I removed all the files, downloaded them a third time and all is working again. Wonderful workflows, thanks a ton for them <3

DigitAll

download with browser, 6,7GB is the full size

DigitAll

Hi. check the realvisxl. many reports are for this incomplete download. https://huggingface.co/alexgenovese/reica_models/resolve/021e192bd744c48a85f8ae1832662e77beb9aac7/realvisxlV40_v40LightningBakedvae.safetensors

Haunted Valkyrie

Hi, thank you for the updated Studio, I have run into a problem I dont understand. When using the Halloween workflow it highlights the Controlnet Model with an error : ControlNetLoader Error while deserializing header: MetadataIncompleteBuffer. I installed the new version from scratch and all downloads completed. The first time I downloaded the models some of them timed out, but I did the downloads again and all appears to have gone fine.

Chintalapati Vatsal

Hey! Great piece of work, really impressed with what you have crfeated. I have been using the character creator quite a bit, but saddened to see that it does not work properly for most prompts. Struggles to get a biped character in the turnaround range. any chance you have prompts guide as well to get consistent good results?

DigitAll

Your feedback is truly motivating, and it encourages us to keep improving the platform. If you have any questions or need further assistance as you explore, feel free to reach out. Thanks again!

Gianluca Amati

I am truly impressed by the project you have created using ComfyUI! Your attention to detail and innovation are evident in every aspect of the work. The interface is not only intuitive and user-friendly, but it also significantly enhances accessibility and productivity. The implemented features not only meet current needs but also anticipate future trends. This project is a game-changer, making complex tasks seamless and efficient. Kudos to the team for delivering such an exceptional and well-designed solution. Keep up the fantastic work! I hope to effectively explore every functionality of this vast, boundless, yet accessible project

Hadi Nooradini

hi add workflow for photoshop plugin in comfy for retoch and makeup direct conrct from photoshop to comfy!

DigitAll

Hi. Download the 24GB v63. Do not download models. Replace the models folder from v41 to v63. Then run The script to download all modells.

Richard Mathews

Hi, I think I'm on update v41, can you advise on the best way to update?

LEO 3D

Everything works, it's very cool, thank you very much for your work. I also have a question: is it possible to upscale in batches?

DigitAll

if you will need help, just ask me.

LEO 3D

First I'll try to install it on my PC, I have an RTX 2060 super, so I hope it will work). But in any case I want to say thank you very very very much, your work is incredible!!!

DigitAll

Yes it includes upscaler. Try to run it on colab. I didn't use colab.

LEO 3D

Hi! Does this include an upscaler? And can I run it on google colab?

grumpy grumpy

Amazing new update!!!! All works!! ty for an amazing product!!

grumpy grumpy

UPdate 53 nice surprise works perfectly even if I did have fat fingers and didn't unpack the zip to path lol... absolutely fantastic workflow!!!!!!!!!!!! Thank you very much!! For others that may be interested I generally unpack directly from within the zip, so I don't unpack to path. Which creates the proper directories for the scripts to work. Be mindful of that and you will have no issues or manually create an update folder and put the files and update.bat in there and all will work as intended. TY to the author for putting me straight.

DigitAll

Thank you

Claude GIROLET

with each update of comfyui studio i'm really amazed by the quality of the work you're doing, i'm very happy to have joined your patreon, money well spent to say the least, congrats.

DigitAll

Thank you very much.

grumpy grumpy

I must say having now worked with your comfyui studio, apart from a couple of small errors which were not your doing but rather authors of various nodes that have updated their nodes without notice or reference to what the changes are eg IPADAPTER PLUS lol.. The guy went and did a full rewrite of his nodes, which broke a lot of workflows. Easy enough to fix but you have to do a little head scratching first. Anyways I have tried many work flows and been a member of a few patreon's where the promise of something wonderful has been made by others who have failed dismally in my opinion. Broken workflows things that don't work, an underlying lack of knowledge as to how comfyui works and the environment it needs to have to work properly. A stroke of Genius to have comfyui work in its own environment outside the influence of the rest of the system!!! All of your workflows work and are very easy to understand, minimum nodes, easy to expand on and to add to and expand and experiment with. It is very obvious that you understand the way comfyui works the inherent problems that can arise and you have made exceptional strides into making it a one touch install. Supplying a batch file to download the appropriate loras controlnets etc etc, so that we the user do not have to look for them or double up on renamed files by various authors and not having to have double triple the amount of files just in case something was different. Congratulations on a well thought out installer and process for the novice and experienced user!! Can't wait to see what you add on next, you really have taken out the hours of troubleshooting one would normally take. So in summary Thank you for your very hard work, people this is not easy to come up with!!!

DigitAll

https://discord.gg/qDaDnSHh

∆KR∆SI∆

I'm sorry I can't find your discord, and I don't have facebook.

∆KR∆SI∆

I will send you a screenshot because neither the lora of Alice Delish that you have provided me works, the character that I get is another one, it is as if there was a lora that is running but it is neither mine that in FORGE if it works nor yours, thanks for your quick response.

DigitAll

Hi. Thank you for using studio. Make Shure the workflow and lora are compatible. Flux workflow must use flux loras. And check the settings in lora node, strength must be 1. If it's not working send me screenshot in private message. Facebook or discord.

∆KR∆SI∆

hello thank you very much for your work had long been looking for something like this to work with comfyUI, I wanted to tell you that I do not work the lora I follow the tutorial but I always get the same character and does not update the loras, I do not know if I have to activate it in some extra way that for my knowledge I'm not doing or I have to reset something in the workflow, sorry for the inconvenience and thank you very much for your work.

DigitAll

If a model download is incomplete, , you must delete that one model. And download it again.

Yiluwidi Dreaming

Awww thank you very much!! Appreciated!!!! Something to help the comic creators amongst us.

Richard mills

Hey there, im running the model downloader and got 5 of the models with one of them having an error. If i stop it now to restart my computer (have to for an unrelated reason) will the model downloader know to pick up where i left it or will it try to replace the files? Thanks!

Zoltán István Bíró

THX!

DigitAll

Wiil not work.

Zoltán István Bíró

I find many interesting Flux.1D checkpoints on Civitai. If I simply copy them into the models/checkpoint folder, will they work? Which ones work and which ones don't?

Zoltán István Bíró

THX, it's working!

DigitAll

https://youtu.be/95FcZxOVjbs?si=XG70Fzpl-CZxct37&t=287

Zoltán István Bíró

I like the update very much, although I can't find a good outpaining in FLUX right now. And I know this, user error, but please: I can't activate the selected workflows. Key combination or switch? What should I do?

SC

Funny you mention that. I just modified the NF4 7 to use GGUF and it seems to be working great. Hopefully we can see some GGUF additions in the near future.

Andrey Lysenko

thanks foe update! Waiting ggufs. Also will be nice to add some of LLM for img2img or upscale workflow.

DigitAll

Write me on discord with screenshots and i will help resolve this. Antivirus blocking the scripts.

DigitAll

Just rename the main folder where the version name is located.

Scubamojo

I do it exactly and precisely as described in your video. I continue to run the Update_Script as Admin. Then there are error messages regarding xcopy and timeout, among other things. The folders workflows, models, user default workflows, pyssssss-workflows have been deleted. It also says Failed to run download script.

Claude GIROLET

it would be nice to have the name of the current version we're using in the top of the interface, for example if the update from 43 to 52 worked we should have comfyUI v52 in the top of the interface, then we'll know where we are

Claude GIROLET

hello, comfy studio is great, would it be possible to add a custom folder where we can automatically store our custom workflows (the ones we derived from your basic workflows), in order to backup them all at once and then reimport it globally after an update.

DigitAll

Please describe the problem with updates. V52 update has issues?

Scubamojo

Hello. I really like your work and ComfyUI Studio. It has real potential. But what really gets on my nerves is that I have already carried out an update exactly according to your instructions for the second time and have ruined the entire installation as a result. That doesn't work, as you describe in your YouTube video! So you should urgently revise your update routine. And it would be nice if it were possible to download the latest version directly and reinstall it. Thank you and keep up the good work (apart from the updates ...)

Mikhail Ershov

how do you do a clean one? download 43 and then upgrade to 52? and new theme in flux in 53?

DigitAll

Yes. Allways will be a clean install after 1 update.

Mikhail Ershov

thanks, but can you also post a clean version - fully v53. for clean installation. without these updates?

DigitAll

Ok

Yiluwidi Dreaming

Please create create comic workflow...??

Andre Prior

When is release?

DigitAll

Today will be released 7 flux dev and 7 flux nf4 workflows. With generation time issues write me in private message to solve it.

Andre Prior

Since the update FLUXInf takes ages.. 1 image 1 hour on a 4060

Andre Prior

will more tutorials follow? yours are awesome

DigitAll

Download the v43 10GB. Unzip in the main folder, run the update script.

Andre Prior

how to update from 41 to 43?

DigitAll

Yes. I make it quick. Next update will be with multi loras.

Raymond Brettschneider

Would it be possible to use a multi LORA loader in those places where single LORA loading is now possible? Sometimes, a mix of LORAs with the right weights is just what's needed to get what you want. And I think a multi LORA loader would slot in just right in the place where there is now a single LORA loader. A multi LORA loader will still allow for loading just a single one for when that's enough. Also, if possible, I'd love to see LORA loading as an option in the 2000px Render + Facefix workflow. I really love that workflow, and it's the one option I'm missing right now :) Thank you very much for all your hard work in developing this set!

Zoltán István Bíró

From what I know so far, the FP16 model would give the best quality. If you could fit that in, that would be great!

DigitAll

Thank you. Now I'm working on flux workflows. 2 days and it will be available. Ok i make similar out painting.

Zoltán István Bíró

Thank you for this beautiful work! I miss two things: one is that you came up with a great solution for outpainting. For example, I am thinking of this: https://www.youtube.com/watch?v=iMxP6sS_nw4 . Since you do not suggest to change the installation, I would be very happy if you could include the outpainting solution you think is good in the ComfyUI Studio installer. Another request: I have exactly the RTX 3090 24GB video card mentioned as an example in ReadMe. And I'm glad that - as you write - you optimized the package for speed AND quality, I would be very happy if you had a Flux workflow version specifically focused on QUALITY. I'll be happy to wait when generating images if I get quality in return! Thank you very much!

DigitAll

The scripts won’t work, but ChatGPT can convert them to a Mac version. The workflows will work as expected. Models must be downloaded manually or with the converted scripts. Open the model script in a text editor; you’ll find the download link and the folder path where the model should be placed. Nodes can be downloaded using the manager.

aziz hilami

Hi, i am on a mac apple M1 Max 64gb memory, i have comfyui installed. and downloading ComfyUI Studio v43 now. I wont get into trouble right?

DigitAll

Good idea. Thanks. Will make it.

Jürgen Mayr

Just an idea.. remove the "press space to continue" in the download batch file. This way it could download overnight on slow connections.

DigitAll

I will update that workflow. And let you know.

DigitAll

Delete the yellow ip adapter node. It will fix that

DigitAll

Discord?

thor42

I'm not on facebook, but i'll try to register there. Seems facebook need some time to let me in.

Shealan Forshaw

Did you find a solution? I have the same problem.

DigitAll

Wich social media do you have to help you. Comment section is not good for help.

Kiriko

Thank you, I don't have facebook... Reactor I will follow the instruction on github page for manual installation, is installed, visual studiois updated, I have 3090TI card, still get error on load the workflow faceswap , thank you

DigitAll

I can help you. Write me on facebook page.

DigitAll

Yes. I can help you. Write me on Facebook

Kiriko

Hello i'm new, just bough the subscription plan. I have problem installing : Reactor, Photomaker, and Instant ID, ComfyUI ask me to try fix, but after restart module will not be updated or fixed, any idea, or suggestion for install manually? thank you :)

Max Mira

does this mean this is not a flux(+lora) workflow yet? I opened it, it's not Flux yet

thor42

Trying the "PRODUCT BG" workflow, but even after redownloading models i always get the following error: 2024-09-23 16:51:18,029 - root - INFO - Starting server 2024-09-23 16:51:18,029 - root - INFO - To see the GUI go to: http://127.0.0.1:8188 2024-09-23 16:51:43,188 - root - INFO - got prompt 2024-09-23 16:51:43,627 - dinov2 - INFO - using MLP layer as FFN 2024-09-23 16:53:16,143 - root - INFO - model weight dtype torch.float16, manual cast: None 2024-09-23 16:53:16,144 - root - INFO - model_type EPS 2024-09-23 16:53:35,857 - root - INFO - Using pytorch attention in VAE 2024-09-23 16:53:35,858 - root - INFO - Using pytorch attention in VAE 2024-09-23 16:53:43,370 - comfyui_segment_anything - WARNING - using extra model: /home/volker/docker/comfy-ui/ComfyUI/models/sams/sam_hq_vit_h.pth 2024-09-23 16:55:17,516 - root - INFO - Requested to load CLIPVisionModelProjection 2024-09-23 16:55:17,516 - root - INFO - Loading 1 new model 2024-09-23 16:55:17,809 - root - INFO - loaded completely 0.0 3522.953369140625 True 2024-09-23 16:55:18,666 - root - ERROR - !!! Exception during processing !!! Error(s) in loading state_dict for Resampler: size mismatch for proj_in.weight: copying a param with shape torch.Size([1280, 1280]) from checkpoint, the shape in current model is torch.Size([1280, 1664]). 2024-09-23 16:55:18,667 - root - ERROR - Traceback (most recent call last): File "/home/volker/docker/comfy-ui/ComfyUI/execution.py", line 317, in execute output_data, output_ui, has_subgraph = get_output_data(obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/volker/docker/comfy-ui/ComfyUI/execution.py", line 192, in get_output_data return_values = _map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/volker/docker/comfy-ui/ComfyUI/execution.py", line 169, in _map_node_over_list process_inputs(input_dict, i) File "/home/volker/docker/comfy-ui/ComfyUI/execution.py", line 158, in process_inputs results.append(getattr(obj, func)(**inputs)) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/volker/docker/comfy-ui/ComfyUI/custom_nodes/ComfyUI_IPAdapter_plus/IPAdapterPlus.py", line 822, in apply_ipadapter work_model, face_image = ipadapter_execute(work_model, ipadapter_model, clip_vision, **ipa_args) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/volker/docker/comfy-ui/ComfyUI/custom_nodes/ComfyUI_IPAdapter_plus/IPAdapterPlus.py", line 441, in ipadapter_execute ipa = IPAdapter( ^^^^^^^^^^ File "/home/volker/docker/comfy-ui/ComfyUI/custom_nodes/ComfyUI_IPAdapter_plus/IPAdapterPlus.py", line 71, in __init__ self.image_proj_model.load_state_dict(ipadapter_model["image_proj"]) File "/home/volker/docker/comfy-ui/comfyui/lib/python3.11/site-packages/torch/nn/modules/module.py", line 2215, in load_state_dict raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format( RuntimeError: Error(s) in loading state_dict for Resampler: size mismatch for proj_in.weight: copying a param with shape torch.Size([1280, 1280]) from checkpoint, the shape in current model is torch.Size([1280, 1664]). 2024-09-23 16:55:18,668 - root - INFO - Prompt executed in 215.43 seconds

Max Mira

with lora or with Swarm UI?

Shubz World

i already did that and everything works fine

Max Mira

Sorry! I have SWARMUI installed and how can I add all this, can you advise?

Max Mira

I'll download it now, is it present inside? Or should I add nodes? If I add manually, will everything work correctly?

DigitAll

ok will make flux lora workflow

Max Mira

Can u upgrade workflows to Flux and Flux+Lora?

Raymond Brettschneider

Thank you very much! I couldn't find that information :-) It's working now.

DigitAll

it must be there. if you updated with the 10GB update. look for it in The 1-Workflows folder. If its there, then run the script : Copy to comfyUI.

Claude GIROLET

since a reinstalle v43 from fresh i can't find some workflows like the 1845 artists, is it possible to import them from previous versions if i decompress v36 ou v41 besides my v43 installation without launching them ?

DigitAll

Hi. If you see Lora nodes disabled, just enable them by right-clicking on the node and clicking on the 'Bypass' option. If any of the workflows do not have a Lora node and you would like to add one, let me know and I will add it.

Raymond Brettschneider

Hi, in the Youtube video I'm seeing workflows with LORA support, in my fresh V43 install I am not seeing these. Is there anything else I need to add for this? Or is LORA now supported in a different method in the regular workflows? I do see some LORA boxes in some workflows, but these are mainly disabled.

Judi Godvliet

Nf4 is a quantized version of Flux isn't? What I meant is most of your great workflows are SXDL workflows. Would be awesome when you are able to update these workflows to Nf4 workflows :)

Claude GIROLET

thanks for your answer, i will try that despite the fact it could be a quiet tedious process, because doing this i could only use one wildcard instead of many in the same prompt, anyway comfy studio is great i have no doubt it will progress in a future. another little question is is possible to set another output directory ? than the default one ? BTW i've checked the app subdirectories and there are no wilcards dubdirectory in it, so for now i think we can't use wildcards in comfyui as we did in fooocus, maybe in a future update they'll allow it.

DigitAll

Test the 1845 artist workflow. Delete the artist list. Paste there any text, each line will be added to your main prompt. But ok will make it.

Claude GIROLET

is it possible to use wildcards in prompt, i'm more used to fooocus where i have a list of text files in a wildcards directory, in each text file there is one parameter per line and when i want to use it as a part of a prompt i type (for exemple) face portrait of a __cats__ and fooocus replace __cats__ with each line in the cats.txt in my wildcards directory, is is possible in comfy studio, and if so could you do a video on how to do it.

DigitAll

Tested the Flux models with controll net. Image quality is poor. Waiting for updates. Nf4 is the best today.

Judi Godvliet

Haha thank you, btw would be great when you could translate the current workflows into Flux based workflows, are you planning this?

DigitAll

yes in the setting: "Graph" switch from Hidden to Linear

Judi Godvliet

Wow, great update. Is it possible to make the node lines visible? I want to adjust some nodes to my own preferences.

21machines

I didn't read all the comments. So I'll ask; Are there any new workflows?

Yiluwidi Dreaming

Curious - any reason on startup it gives this 'Python was not found; run without arguments to install from the Microsoft Store, or disable this shortcut from Settings > Manage App Execution Aliases. git version 2.45.2.windows.1'

SoFiB34tz

before i tried to install v43 from scratch, just now i reinstall v36 then i will try to update to v43, if I encounter the same problem, I will contact you on facebook, ok?

DigitAll

No problem.

Yiluwidi Dreaming

Nope. Seems I will have to Copy models across from my Hard Drive Install. Bit of a drag man...oh well might be a chance to clean my current models anyhoo... The good thing is though the workflows are all functional (as long as you have the appropriate models) not like some hit and miss workflows on open art. It appears there may be some flexibility in using your own workflows and models....or is that a problem??

DigitAll

can you write me on facebook to help you?

SoFiB34tz

there is a checkmark on desktop development with C++ so i think it's ok

DigitAll

Check it. Do you have the right workload. "Desktop development tools"

SoFiB34tz

I installed the C++ packages in addition because I did only the python ones when initially installed Visual Studio

DigitAll

Did you install the right workload for Visual Studio?

SoFiB34tz

I had Visual Studio installed already before.

DigitAll

The models downloaded you can use from the models folder. Cut and paste to the new v43.

DigitAll

Unzip a new studio v43.

DigitAll

Install visual Studio. Like in the video for the installation. Will be fixed.

SoFiB34tz

Hi, I also unzipped the file, ran the update all models and all good, but when I start -RUN-VENV and switch to any of the ID workflows I get the error. Missing node: ApplyinstantID, InstantIDFaceanalysis, and InstantIDmodeLoader. I then tried different solutions from the github of InstantID, Reactor and InsightFace, also did some pip downgrades for example numpy, now my complete installation seems broken. Is there any possibility to bring it back to work?

DigitAll

Can you write me on Facebook?

Yiluwidi Dreaming

I already have comfy installed on C drive ** Python version: 3.11.9 This is what happens when i try to run studio .... on external drive. Python was not found; run without arguments to install from the Microsoft Store, or disable this shortcut from Settings > Manage App Execution Aliases. git version 2.45.2.windows.1 [START] Security scan [DONE] Security scan ## ComfyUI-Manager: installing dependencies done. ** ComfyUI startup time: 2024-09-22 15:38:36.683513 ** Platform: Windows ** Python version: 3.11.8 (tags/v3.11.8:db85d51, Feb 6 2024, 22:03:32) [MSC v.1937 64 bit (AMD64)] ** Python executable: D:\ComFy_Studio\App\python_embeded\python.exe ** ComfyUI Path: D:\ComFy_Studio\App\ComfyUI ** Log path: D:\ComFy_Studio\App\comfyui.log Prestartup times for custom nodes: 0.0 seconds: D:\ComFy_Studio\App\ComfyUI\custom_nodes\rgthree-comfy 0.0 seconds: D:\ComFy_Studio\App\ComfyUI\custom_nodes\ComfyUI-Easy-Use 1.3 seconds: D:\ComFy_Studio\App\ComfyUI\custom_nodes\ComfyUI-Manager Total VRAM 16380 MB, total RAM 32649 MB pytorch version: 2.4.1+cu121 Set vram state to: NORMAL_VRAM Device: cuda:0 NVIDIA GeForce RTX 4060 Ti : cudaMallocAsync Using pytorch cross attention [Prompt Server] web root: D:\ComFy_Studio\App\ComfyUI\web D:\ComFy_Studio\App\python_embeded\Lib\site-packages\kornia\feature\lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead. @torch.cuda.amp.custom_fwd(cast_inputs=torch.float32) [Crystools INFO] Crystools version: 1.16.6 [Crystools INFO] CPU: AMD Ryzen 5 5600 6-Core Processor - Arch: AMD64 - OS: Windows 10 [Crystools INFO] Pynvml (Nvidia) initialized.

DigitAll

Downloading with scripts sometimes results in incomplete downloads. Delete the faulty model that the workflow is using and manually download it via the browser. You can find the link to the model in the script.

DigitAll

Downloading with scripts sometimes results in incomplete downloads. Delete the faulty model that the workflow is using and manually download it via the browser. You can find the link to the model in the script.

Yiluwidi Dreaming

update finished - same issue - BTW this IS the screen shot!! Happens when try to download models... D:\ComFy_Studio\update-v43\python_embeded\share\man\man1\svg2pdf.1 D:\ComFy_Studio\update-v43\python_embeded\share\man\man1\ttx.1 D:\ComFy_Studio\update-v43\python_embeded\tests\data\scotland_house.png D:\ComFy_Studio\update-v43\python_embeded\tests\data\scotland_pitie.png D:\ComFy_Studio\update-v43\python_embeded\tests\data\scotland_plain.png 66199 File(s) copied -------------------------------------------------- Note: If a model does not work, it means the download was interrupted during the process, so you need to delete it and download it again. ------------------- Running D:\ComFy_Studio\5-MODELS\Clipvision for ipadapter.bat... Downloading clipvision-for-ip-adapter.safetensors to D:\ComFy_Studio\5-MODELS\..\App\ComfyUI\models\clip_vision... % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0 curl: (35) schannel: next InitializeSecurityContext failed: CRYPT_E_NO_REVOCATION_CHECK (0x80092012) - The revocation function was unable to check revocation for the certificate. Download completed. Press any key to continue . . . Running D:\ComFy_Studio\5-MODELS\diffusers_xl_canny_full.bat... Downloading diffusers_xl_canny_full.safetensors to D:\ComFy_Studio\5-MODELS\..\App\ComfyUI\models\controlnet... % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0 curl: (35) schannel: next InitializeSecurityContext failed: CRYPT_E_NO_REVOCATION_CHECK (0x80092012) - The revocation function was unable to check revocation for the certificate. Download completed. Press any key to continue . . .

Shealan Forshaw

I am running on a 4090

Shealan Forshaw

I installed your updated #43 from scratch and redownloaded all the models. They all seem to load successfully. However, when running some workflows (eg: Product background replacement) I get the following error: Error(s) in loading state_dict for Resampler: size mismatch for proj_in.weight: copying a param with shape torch.Size([1280, 1280]) from checkpoint, the shape in current model is torch.Size([1280, 1664]).

Yiluwidi Dreaming

Ok. Onnit.

DigitAll

Fixed this issues. They're fixed now. Download available.

DigitAll

Fixed. Download available

Yiluwidi Dreaming

Looks like might be covered as a fix in new update?? Give a little hint - am I right??

DigitAll

check wich model is with error, download it manually from huggingface. google it. or try with the script. delete the old model before downloading

MHI Ir

Hi i installed comfyui-studiob But to use their work follows, I encounter this message. Error while deserializing header: MetadataIncompleteBuffer If it is due to the incomplete download of the models. Please download all the packages at once.

DigitAll

The workflow already generates PNG images with transparent backgrounds in the output folder. Could you clarify if there's something else you're looking for?

Oviyan

yes i used it, i want to generate png images directly

DigitAll

updated the v43 archive. now in the REMOVE BG workflow is a batch option.

DigitAll

ok

Oviyan

Can you add a layerdiffusion with batch

DigitAll

Uninstall those 2 nodes. I updated workflows 10 minutes ago. To work without them.

Shealan Forshaw

I ran the #41 update which worked. Then the #43 update but 2 modules fail to import when Comfy is started. Any ideas what I can try to fix? 0.0 seconds (IMPORT FAILED): D:\ComfyUI Studio\App\ComfyUI\custom_nodes\ComfyUI-GGUF 0.0 seconds (IMPORT FAILED): D:\ComfyUI Studio\App\ComfyUI\custom_nodes\ComfyUI-Crystools

DigitAll

Wat do you have to send to send screenshots?

Yiluwidi Dreaming

I dont have FB - give me the fix pls

DigitAll

write me on facebook please. i will help

Yiluwidi Dreaming

Everything DOES not work perfectly as I got the following when attempting to install models: curl: (35) schannel: next InitializeSecurityContext failed: CRYPT_E_NO_REVOCATION_CHECK (0x80092012) - The revocation function was unable to check revocation for the certificate. Download completed. Press any key to continue . . . I already have another installation on C drive. Comfy Studio installed on External Hard Drive...... update41 loaded

Patrick Nielsen

Thanks it works perfect... ☺️☺️

Andre Prior

I only have 56 workflows

DigitAll

Yes. You can use the workflows there. Scripts will not work.

Chris "Broadway" Romero

Hello! Would you recommend installing this in a comfyui_windows_portable setup, I already use many custom nodes, animatediff etc... or should I install separately in its own directory?

DigitAll

Yes. Just use the Cfg, Steps, Sampler, and Scheduler settings for your Checkpoint.

Stacy Burnette

Can I add other checkpoint models in this, I have some favorite I always use

DigitAll

Run the Copy to comfyui script in the 1-WORKFLOWS folder, then press the refresh button in ComfyUI

Patrick Nielsen

I cant seem to find the 1845 workflow after updating?

DigitAll

Check the model. Maybe incomplete download. Delete and download again

Stacy Burnette

seein this error when trying to run CheckpointLoaderSimple Error while deserializing header: MetadataIncompleteBufferCheckpointLoaderSimple Error while deserializing header: MetadataIncompleteBuffer how can I fix

DigitAll

Hi. No it will not cause error to other comfyui. just run the "Download all models" script, and launch with the Run VENV script.

guni

hi just join this patreon and finish download the pack. Is there any installation guide or installation video for this? I am currently have exisiting comfyui installed, will this cause an error to my existing installation?

Richard Mathews

Sorted it with the install of visual studio, thanks.

DigitAll

install Visual Studio 2022 and choose the "Desktop Development with C++" workload.

Richard Mathews

Hi, i haven't updated, just unpacked the rar and downloaded the models then run the bat file: -RUN-VENV-.bat

DigitAll

did it worked before update?

Richard Mathews

Hi question. Unzip the file, ran the update all models and all good, but when I start -RUN-VENV and switch to any of the ID workflows I get an error. Missing node: ApplyinstantID, InstantIDFaceanalysis, and InstantIDmodeLoader. Do I update comfu or look for missing nodes? I notice that the import fails when looking for missing nodes.

DigitAll

The Mac's video card cannot handle the load required for ComfyUI.

Alex Fetter

Is there also a version for Mac?

Elwaves

Flux with multiple loras is what I'm waiting for. I already have it in another workflow/install, so no rush for me but it would be great for it to be part of this so I can ditch the other.

AI_Guru

Thank you

thor42

Here we use Windows (bat files ...) and I'm on Debian. Therefore it took some time to check the correct Ids of the custom nodes used here. The CSV-Loader (333) uses Windows path separators, after changed those inside csv_loader.py throwing the styles.csv into the correct folder, it seems to run smoothly. This is the list of custom nodes that are used, others were disabled to run this. 1 ComfyUI-Manager 2 ComfyUI Impact Pack 3 ComfyUI Inspire Pack 8 ComfyUI's ControlNet Auxiliary Preprocessors 64 pythongosssss/ComfyUI-Custom-Scripts 71 UltimateSDUpscale 81 Comfyroll Studio 99 ComfyUI_IPAdapter_plus 100 ComfyUI InstantID (Native Support) 138 ReActor Node for ComfyUI 147 Styles CSV Loader Extension for ComfyUI 176 rgthree's ComfyUI Nodes 182 Save Image with Generation Metadata 215 ComfyUi Nodes for External Tooling 216 ComfyUI Inpaint Nodes 253 KJNodes for ComfyUI 333 ComfyUI-CSV-Loader 431 ComfyUI Easy Use 694 ComfyUI-layerdiffuse 814 ComfyUI-AutoCropFaces 1172 ComfyUI-tbox 1196 ComfyUI-AdvancedLivePortrait Nice experience so far. Keep up with it.

AI_Guru

I use my own setup i use style.csv in ComfyUI/styles.csv

DigitAll

Did you download the 20gb pack?

DigitAll

If you're using the 20GB pack, then it's already in place. If you're downloading only the workflows and using your own ComfyUI setup, place the style.csv file in the \ComfyUI\ folder.

AI_Guru

Where should i find input prompts

AI_Guru

Where should i paste the style.csv in comfyui

DigitAll

1. DOWNOLAD ALL MODELS. wait until its downloading. 2. RUN VENV This start the ComfyUI

moti beny

is there an explanation about what bat file to activate and what second ?

DigitAll

If the error appears, message me on Facebook, and I’ll help you.

DigitAll

Just double click on RUN VENV

Elegost

Now I get this when trying to launch it: Windows cannot find 'D:\AI\ComfyUI VENV\App\-RUN VENV-.bat'. Make sure you typed the name correctly, then try again.

DigitAll

Only one

Value Store Zone

do i download all three links or just one and extract it?

Elegost

Thanks! Yeah only Winrar worked. The rest crapped out half way through.

DigitAll

winrar app. 5 minute max

Elegost

I'm having trouble unzipping the files. How long does it take for you to unzip? What software do you use?

Shubz World

supported little much as i could,

Zakaria Nada

I use it on a daily basis, really handy

Zakaria Nada

I agree, more FLUX workflows. You can do the lora training workflow, Flux-dev-lora and multiple loras, Flux controlNet... These are things that not so many do and it would A+ for your project

Judi Godvliet

Unbelievable you are really a Comfy Wizard! Btw, could you add more FLUX workflows in version 37 :)

O.M.3

and i would like to support your work, a created a new workflow "ControlNet union sdxl" it's really good, i can re-create i character in different style, if want it just let me know. "that is a small gift for your awesome work this an example https://youtu.be/PhrScECoCL0

O.M.3

Hi, I want to thank you for this awesome workflows I've tried it and it is really give me a good result, thank you very much. And I have a suggestion for a new workflow in ID+POSE wrokflow, there is a note called depth anything. switch the node with other like canny or line art etc, that will be awesome

Tobe2d

I understand for the workflows. But I am talking about comfyui itself and the addons inside it.

DigitAll

New workflows will be available via new links I provide. They will include an updated ComfyUI app and new workflows.

Tobe2d

Thanks for the reply, so how we do the updates?

DigitAll

Please do not update ComfyUI manually. Every pack is already updated.

Tobe2d

i've tested ComfyUI Studio and it is really nice! Thank you for putting this togather.. Question please, Is it okay to update comfy and nodes to the latest? or it has to stay as it is?

Zom Boy

Can you please do a tutorial on how to run the 56GB file on a virtual GPU such as Vast.ai? Not all of us have powerful computers to run it locally :(

DigitAll

Styles.csv must be in the root ComfyUI folder. Where the "main.py" is.

Richard Mathews

Hi, I received an update on my local PC regarding the styles and all working fine, but I also use a cloud PC called MimicPC. When I upload the styles file to the directory containing the models, the node doesn't recognize it. I suspect there might be a different file structure. Is it possible to edit the styles CSV loader node to point to a new location? here is a screen shot: https://www.awesomescreenshot.com/image/50393345?key=67fbe5e12e400c304fb45486bb0b8236

Mikhail Ershov

we're very much looking forward to it. Still need to mix the imag with another image (ip adapter) + canny (keep recognizability of facial features) + Pose (for consistent character sheet)

Judi Godvliet

Awesome, thanks for all the work!

DigitAll

Yes, ControlNet workflows will be available soon.

Gabriel

Any plans to add img2img workflows?