Download ComfyUI Studio

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

DigitAll

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

Hugh Milstein

Read this post.

DigitAll

how do i downlad the package?

hitesh parmar

RTX 3080 10gb

Stephen Orsatti

Yes

DigitAll

Rtx 3090?

Laurent Cresson

and for runpod?

red baron

not for videos. only for images

DigitAll

rtx 5060?

jonh doe

are you sure?

Laurent Cresson

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

DigitAll

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

Hans

Hi. Tomorrow

DigitAll

Hi when does complete installation files v305 Will be downloadable?

Laurent Cresson

For image generation yes. For video no

DigitAll

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

ai.playhardgopro

Yes every week will be more.

DigitAll

Update 288 All models download size actually is 512GB

Andres Lentijo

do you use discord?

DigitAll

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.

Laurent Cresson

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?

Kara Concepcion

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

DigitAll

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

john Greene

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

Serdar Polat

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

Laurent Cresson

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.

Jose M. Aranda

Go to discord "download studio" channel

DigitAll

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

Greg Waters

Ok will send you in message

DigitAll

Hi discord blocked in Turkey can you send other way

Ulaş Aksan

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

DigitAll

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

Unbound Logic

Write me on discord

DigitAll

Hey Alex I cannot find the update link in discord

Roberto Strempler

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

Charles Phillips

Just renamed it from 278. I deleted unused workflows

DigitAll

v256 ? is that a typo

m ab

did you connect patreon account with discord account?

DigitAll

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

Lukas Neumann

I dint see u in my discord at all and really need to update the studio its outdated and won't update the nodes

Chris McKellen

Connected Patreon with discord?

DigitAll

i cant find you in my discord

Chris McKellen

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

jonathan sanchez

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

DigitAll

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

michel sneyers

256 is latest

DigitAll

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

Trader McBook

This is the latest

DigitAll

Is this Revision 258 OR Revision 276?

M. Ray White

not supported

DigitAll

How well does this run on runcomfy or think diffusion?

Austin Wilson

Did you connect Patreon with discord? In Patreon settings

DigitAll

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

Desmond Lee

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

DigitAll

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.

LVNDSCAPE

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

DigitAll

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 . . .

Rayeta76

Yes.

DigitAll

we have a downloadable pdf with the changes

Rayeta76

Yes. I renamed it to Depthmap 4

DigitAll

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

Bill P

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

Supergonzo2011

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

DigitAll

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

Wills_VFX

Nevermind I got the readme file

Gregory Contreras

Which workflow you recommend me for Background Removal?

Gregory Contreras

use pinochio app to install fluxgym

DigitAll

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?

Мария Павлова

Thank you very much for the new update!

Alexey Trusov

Thanks. Enjoy

DigitAll

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

Jose M. Aranda

build is comlex to use it on other device

DigitAll

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.

Владимир Кунаков

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

Art Szabo

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

marwa velan

yeees please workflow with upscale video <3

Nik

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

Jose M. Aranda

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

m ab

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

Mohammed Allootah

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

DigitAll

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 ""

Mohammed Allootah

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 ..

Kerem Copcu

No need.

DigitAll

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

O.M.3

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

DigitAll

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" (

red lateral

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.

LVNDSCAPE

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

DigitAll

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

Laurent Cresson

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

DigitAll

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.

Wolfgang D.

Write me on discord

DigitAll

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?

Vahan Terterian

Yes

DigitAll

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

Steffen Kirkegaard

I have a generally poor output quality

Many Wayne

voice clone you can put the brazilian portuguese voice

Paulo Chiovatto

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?

M. Ray White

ok

DigitAll

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

Bolek Lolek

Send me screenshot on discord

DigitAll

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')

Nicolas Calabrese

Just close this message. Its not error

DigitAll

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

denni dondi

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.

Sai Xiong

thanks will check it and fix it.

DigitAll

Hi! V229 Wan video nodes doesn't work

Alex Qwerty

what error?

DigitAll

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

Alfonsozm

Send me message on discord i fix it

DigitAll

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

WCMT2020

Diffusers MV Sampler is not fixed on Workflow Character V2

Bruno Solano

Download the update 16MB. It will fix it

DigitAll

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

denni dondi

Send me message on discord to help you

DigitAll

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

Wupi

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

DigitAll

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!

VirtualBite

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

DigitAll

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)

Manuel Leiton

fixed

DigitAll

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

Nik

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.

Fernado Landín Urbina

Write me on discord

DigitAll

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

Fernando Giovanini

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

Tyler

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

Iván Holguín

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.

Suresh K. Volam

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."

RMonster

error wan Video, Missing node type CFGZeroStar WanFunControlToVideo

Rodrigo Mariano

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

Javi Sanchi

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?

Bill Hancock

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

DigitAll

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.

Laurent

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

DigitAll

link for updates not working

Bill P

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

DigitAll

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?

Stanislav Stoyanov

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

Wari777

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.

Deluxe Music Selection

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.

bruno davi

ahh ok haha, that must be it :)

Cris

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

Johnathan B

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

Cris

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

Johnathan B

cant be done

DigitAll

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

William Van Leer

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

Wolfgang D.

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.

Gamearc

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

John

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.

Mike Chandler

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

4DXP

The models you can use from the old

DigitAll

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

Andy Beck

218 update sir

Banchert VANTHA

Its ok.

DigitAll

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.

Wolfgang D.

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

DigitAll

gersongabrielgm

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.

Gorka Soria

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

R P

Try the reserve link, that one worked for me

Alfonso Marquez

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

TacImpulse

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…👍

AirborneSapper708 RoughTerrain12B

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

Gurmeet Singh

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

DigitAll

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

Steffen Kirkegaard

oh thank you

Deluxe Music Selection

On discord you can download it

DigitAll

Its in update v215 archive

DigitAll

would be nice only to download the workflow :)

Deluxe Music Selection

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

JOHN VOLTAIRE GARCIA

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

Desmond Lee

No. Must update 205

DigitAll

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

Desmond Lee

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.

Games_Graphics

you must send me screenshots on discord or facebook

DigitAll

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

kamil

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

주현 한

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

DigitAll

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.

bruno davi

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

Alfonso Marquez

ok

DigitAll

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

Many Wayne

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

DigitAll

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

Gorka Soria

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

DigitAll

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

Thinh Pham Duc

THANKS BRO

Abdullah Elsayed Bakry Tolba

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

DigitAll

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

Abdullah Elsayed Bakry Tolba

Screenshot full error. On discord or Facebook send me

DigitAll

Apply Whisper error in audio 3 folder [WinError 2]

Roman Yuferev

yes download links are in the post

DigitAll

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

Games_Graphics

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.

LVNDSCAPE

none of the links work on 205

Gravity

links active now

DigitAll

links active

DigitAll

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

DigitAll

Hello where are the links

Frédéric

Google limit achieved. Try later.

DigitAll

Links are not working!

Jose G Garza

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

Kevin O'Malley

Yes 6.7gb. i forget to edit it

DigitAll

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

Serkan

Maybe something is blocking the scripts.

DigitAll

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

Thinh Pham Duc

Hi can you add, SkyReels and hunyuan image2video please?

Serkan

Send screenshots on discord

DigitAll

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

Олег

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

Manuel Nixon

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

Олег

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?

William Van Leer

Yes.

DigitAll

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

Gamearc

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

Олег

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.

gersongabrielgm

Extract 9 archive files with WinRAR app.

DigitAll

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.

gersongabrielgm

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

Олег

Can you help me about it please for Runpod?

Marie

Yes. You can use any lora

DigitAll

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.

William Van Leer

Delete spaces in folder names. D:\comfy ui

DigitAll

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

cessare cessare

Also locally

Ivan Ruiz

I am using it on Run Pod :/

Ivan Ruiz

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.

Dazaeth Garduño

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

DigitAll

William Van Leer

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

Aurimas Gelumbauskas

Yes. Watch the lora video on my you tube

DigitAll

Do you have a workflow for creating digital influencers?

Dazaeth Garduño

When can you update it to work with the 5090?

Khoi Dang

no

DigitAll

Does this work with the Nvidia rtx 5090 / 5080?

Khoi Dang

To know i must see workflow screenshot

DigitAll

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 . . .

Steffen Kirkegaard

Hi. No.

DigitAll

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

Marie

Under the update download link is a youtube link

DigitAll

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."

Jeff Merilo

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

DigitAll

Send me screenshots on discord

DigitAll

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

Desmond Lee

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"

Euge If

hi. yes

DigitAll

Hi! Do you have a workflow with Florence 2?

Dazaeth Garduño

Sent you friend request

Gamearc

How did you resolve it?

Gamearc

Send me workflow screenshot on discord

DigitAll

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'

Gamearc

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

DigitAll

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

Gamearc

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

John McGarrity

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

Jessica Landa

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

Skubz

App\comfyui\models\checkpoints

DigitAll

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

Skubz

Check the models in folder

DigitAll

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

Skubz

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.

DigitAll

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.

Wolfgang D.

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?

Miriam Kassandropolis

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

DigitAll

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?

Arbaaz Nasser

Yes. Now use WinRAR to extract.

DigitAll

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.

Arbaaz Nasser

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?

Soundlufs

Thanks. The Flux upscale is now very good.

DigitAll

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!!

Soundlufs

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

Anwar Hussain

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

DigitAll

I've sent a friend request

Anwar Hussain

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?

MarcAndreMueller

Discord to chat. My username is comfyuistudio

DigitAll

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

Anwar Hussain

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

DigitAll

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

Anwar Hussain

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?

bruno davi

Uninstall latent sync. Only wrapper must be installed

DigitAll

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

Anwar Hussain

Unable to join discord,

Alex Qwerty

Try install fluxgym with pinokio app

DigitAll

@comfyUI Studio ? Need help ASAP

Damien Watkins

send me screenshot in private or facebook i need to see

DigitAll

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

Art Cave

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.

Diego Cavali

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.. "

Damien Watkins

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 ?

Alexandre Massart

T2V tutorial. Any image to video workflow?

Nemesis_Film

Uninstall and reinstall

DigitAll

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

Michael Blendow

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

Nemesis_Film

V184 not active anymore. V190 is active

DigitAll

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

RichardRhal

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??

Bill P

I can help on discord or Facebook

DigitAll

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

Filipe Santos

screensot please on discord

DigitAll

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

Filipe Santos

Install latent sync node from manager, then uninstall it

DigitAll

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'"

종현 한

he literally says use Winrar only..

Nemesis_Film

Yes. There are important files.

DigitAll

I tried it, it didn't work.

Kantemir Maykop

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

Chris Carlson

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

Tyler

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

DigitAll

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.

Chris Carlson

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

DigitAll

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.

Rafael Luna

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

Nemesis_Film

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

Tobe2d

ok will do a detailed video

DigitAll

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.

Relatos para escuchar

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

DigitAll

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?

M. Ray White

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.

Relatos para escuchar

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.

DigitAll

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?"

Relatos para escuchar

Yes

DigitAll

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

Richard Mathews

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

DigitAll

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?

Relatos para escuchar

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

DigitAll

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.

John McGarrity

It allows but not recommend

DigitAll

The name must be original. Check file names

DigitAll

Nemesis_Film

Nemesis_Film

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

DigitAll

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?

Kantemir Maykop

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

Daniela Vergara

Download the v185 workflow update again. I added languagees.

DigitAll

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

DigitAll

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

Daniela Vergara

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

JOHN VOLTAIRE GARCIA

Send me screenshot of the ffmpeg\bin folder

DigitAll

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

JOHN VOLTAIRE GARCIA

sent on discord thanks

John McGarrity

Find me on discord or Facebook to send me private message

DigitAll

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

John McGarrity

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

DigitAll

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

JOHN VOLTAIRE GARCIA

Ok so update don't work. Understand

aprisma

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

DigitAll

Thanks a lot! Problem fixed!

Marian Cepeha

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

aprisma

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

DigitAll

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!

Marian Cepeha

There is one with 3 different methods in v185.

Matt Morton

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

Manish Jain

Added link 2

DigitAll

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

Manish Jain

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.

Wari777

Hi. Send me screenshot of the workflow and error

DigitAll

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

John McGarrity

Amazing.... Thank you for the Watermark Workflow

Deluxe Music Selection

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?

Lazlo

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

Đặng Văn Cường

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!

Ezequiel Casas

Thanks. Enjoy)

DigitAll

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

Alex Young

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

Deluxe Music Selection

Detailed install description fixed.

DigitAll

Thx for your release workflow.

LVNDSCAPE

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

DigitAll

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 ?

LVNDSCAPE

Lora trainer app

DigitAll

What is FluxGym install.zip?

Lazlo

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

gary

Add friend request. Username is comfyuistudio

DigitAll

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

silfras joeld

Write me on discord or Facebook

DigitAll

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

silfras joeld

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

DigitAll

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

marek kazimierz

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

Anthony

Download it manually. The flux model

DigitAll

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?

ruben sebban

Send me screenshot of root folder. And update folder

DigitAll

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 //

종현 한

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!

에밀

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

DigitAll

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?

에밀

ok will try that. Thank you.

Silb3erbart 1

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

DigitAll

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

Silb3erbart 1

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

DigitAll

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 ?

Silb3erbart 1

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

DigitAll

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.

aprisma

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

DigitAll

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

alex47me

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

João C R Fonseca

Which one?

DigitAll

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?

João C R Fonseca

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?

aprisma

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.

Ron Harris-White II

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

DigitAll

Yes, I am

Ron Harris-White II

Hi. You using windows?

DigitAll

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

Ron Harris-White II

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 :)

LVNDSCAPE

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

DigitAll

Write me on discord or Facebook i will help

DigitAll

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.

LVNDSCAPE

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

DigitAll

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.

LVNDSCAPE

Hi. Your internet connection is stable and fast?

DigitAll

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

MoGrafix

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.

DigitAll

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

aprisma

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

DigitAll

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.

aprisma

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?

Andre Silva

Show me screenshots

DigitAll

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

Work Official

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

Claude GIROLET

It's not possible. Wich extractor? WinRAR?

DigitAll

The extractor is asking for a password

Work Official

Write me directly in private

DigitAll

I HAVE THE SAME PROBLEM - V159

Bold

Must see screenshots. Send me on Facebook or discord

DigitAll

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

Daniela Vergara

ok. thank you i will try it

O.M.3

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

DigitAll

https://ibb.co/4Sb3fqS

O.M.3

https://ibb.co/RDDhv3g

O.M.3

Send me screenshot of the models\vae folder.

DigitAll

yes, i tried 2 times

O.M.3

It's not working?

DigitAll

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"

O.M.3

THANKS IT WORKED

Shai Bitton

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.

DigitAll

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"

Shai Bitton

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.

LVNDSCAPE

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

DigitAll

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 :)

LVNDSCAPE

Link 2 added.

DigitAll

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

David Fischer

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.

DigitAll

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

LVNDSCAPE

Yes. Yes. Yes

DigitAll

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?

Laga Mahesa

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!

Manuel Fernandez

Download the model with the scripts.

DigitAll

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 []

Manuel Fernandez

Will make. If you need help write me on discord

DigitAll

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.

Marian Cepeha

Run download all models.

DigitAll

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

Marian Cepeha

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.

Neil Biondich

Write me on discord or Facebook i help you install it

DigitAll

Yes. Use 11.8

DigitAll

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

Frederic Chauveau

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

Charles Phillips

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

DigitAll

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"

Frederic Chauveau

Make a back up. And try it

DigitAll

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

guni

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

M. Ray White

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

DigitAll

what are the changes on 150v?

Miriyala Praveen Kumar

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

DigitAll

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

Muhammad Hussain

Oooh I will try soon, thank you

Madara Uchiha

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

Hussein

It will be in the next post

DigitAll

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.

Imad Elorf

added.

DigitAll

Yes

DigitAll

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

Madara Uchiha

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

Milsen Ace

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

DigitAll

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

Milsen Ace

There is a link in the white node.

DigitAll

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.

Manuel Nixon

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

guni

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

DigitAll

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

guni

with default settings and images?

DigitAll

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

guni

Almost same

DigitAll

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 ?

Hussein

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

DigitAll

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?

Hussein

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

guni

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

DigitAll

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?

guni

Great! Thank you

John Elder

Today will make it.

DigitAll

Do you gave a server on Discord?

John Elder

That fixed it thank you

RudyVRGuy

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 :)

Hussein

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

DigitAll

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']

RudyVRGuy

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.

Imad Elorf

Once again great cinema Alex!

Albert Bardewyck

Which workflow?

DigitAll

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

Imad Elorf

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

grumpy grumpy

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.

Jeramy Udy

142 wow ty

Nobodys Hero

yes. download the update file v142.

DigitAll

Is v96 released ?

Nobodys Hero

perplexity ai works wonders too

justin ruscheinski

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

grumpy grumpy

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!

Nobodys Hero

Tomorrow will release the v96. Lots of new workflows

DigitAll

whats new in v88

Nobodys Hero

Yes

DigitAll

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?

Richard Mathews

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

Dhizzylab Gon

maybe you have renamed some parts, or incomplete download.

DigitAll

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

DigitAll

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.

Dhizzylab Gon

Ask me first if you need help. )

DigitAll

Thanks you saved me a lot of headaches 🙌😊

Soufiane Zahir

Write me on Facebook or discord to see your files.

DigitAll

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

wilBr

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

DigitAll

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

Soufiane Zahir

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

DigitAll

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

wilBr

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

Tyler

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

DigitAll

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...

Tyler

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.

RuggedPineapple

Is there a way using this workflows in mimicpc comfyui?

Mert Durmazer

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

Claude GIROLET

I have been making symlinks for things like the wildcards.

Jeramy Udy

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

Claude GIROLET

hi chat gpt can quick convert them for you.

DigitAll

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.

jc

write me on discord or facebook, send me screenshots

DigitAll

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.

Ora Davis

Write me on discord or Facebook i wii help you

DigitAll

how to update v63 to v 86?

Hakan Bulut

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

Thái Xuân Anh

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

Thorfs

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

DigitAll

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

Thorfs

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

DigitAll

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

Thorfs

? Describe please

DigitAll

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

Thái Xuân Anh

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

Kusacubari

ok. I attached to this post.

DigitAll

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?

oscar DCDC

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...

grumpy grumpy

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

DigitAll

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

Tobe2d

No. Not for now.

DigitAll

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?

guni

In the input folder

DigitAll

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

guni

Fantastic!

Ora Davis

how long is the video?

DigitAll

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

guni

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

DigitAll

send me screenshots on discord. Username: comfyuistudio

DigitAll

Link 3 doesn't work ! Downloadlimit

Peter Uwe Reif

"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?

fatih uzun

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

Peter Uwe Reif

Thank You! THat worked

Chintalapati Vatsal

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

DigitAll

Thanks ! It works ! :)

GhØst•WØlf

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

Roman Yuferev

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

DigitAll

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

ricardo rizzutti

i'm working on link 3.

DigitAll

i'm working on link 3.

DigitAll

ok. i'm working on link 3.

DigitAll

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

Chintalapati Vatsal

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

GhØst•WØlf

The same problem, download limit on both links.

Peter Uwe Reif

Succes downloaded via link 2. Tqvm

Rul Cspm

Try Link 2

DigitAll

you have a download limit from google drive, fix it

Денис Балацкий

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

DigitAll

how update v63 to 86 ?

Hadi Nooradini

me2

Hadi Nooradini

same

TheAnim8tor

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

Rul Cspm

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

Tobe2d

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

DigitAll

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

Tobe2d

Wait a little. New update will be released today.

DigitAll

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

red baron

thanks thanks

ricardo rizzutti

Hi. Thank you. Uninstall cuda software.

DigitAll

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."

ricardo rizzutti

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.

DigitAll

Cute you workflow isn't loading...

Chris Ife

Thank you. Never delete models folder.

DigitAll

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

Haunted Valkyrie

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

DigitAll

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.

Haunted Valkyrie

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?

Chintalapati Vatsal

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!

DigitAll

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

Gianluca Amati

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

Hadi Nooradini

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.

DigitAll

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

Richard Mathews

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?

LEO 3D

if you will need help, just ask me.

DigitAll

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!!!

LEO 3D

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

DigitAll

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

LEO 3D

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.

grumpy grumpy

Thank you

DigitAll

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.

Claude GIROLET

Thank you very much.

DigitAll

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!!!

grumpy grumpy

https://discord.gg/qDaDnSHh

DigitAll

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.

∆KR∆SI∆

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.

DigitAll

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.

∆KR∆SI∆

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

DigitAll

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

Yiluwidi Dreaming

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!

Richard mills

THX!

Zoltán István Bíró

Wiil not work.

DigitAll

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!

Zoltán István Bíró

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

DigitAll

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?

Zoltán István Bíró

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.

SC

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

Andrey Lysenko

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.

DigitAll

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.

Scubamojo

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.

Claude GIROLET

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

DigitAll

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 ...)

Scubamojo

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

Mikhail Ershov

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

DigitAll

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

Mikhail Ershov

Ok

DigitAll

Please create create comic workflow...??

Yiluwidi Dreaming

When is release?

Andre Prior

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

DigitAll

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

Andre Prior

will more tutorials follow? yours are awesome

Andre Prior

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

DigitAll

how to update from 41 to 43?

Andre Prior

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

DigitAll

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!

Raymond Brettschneider

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

Zoltán István Bíró

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

DigitAll

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!

Zoltán István Bíró

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.

DigitAll

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?

aziz hilami

Good idea. Thanks. Will make it.

DigitAll

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

Jürgen Mayr

I will update that workflow. And let you know.

DigitAll

Delete the yellow ip adapter node. It will fix that

DigitAll

Discord?

DigitAll

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

thor42

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

Shealan Forshaw

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

DigitAll

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

Kiriko

I can help you. Write me on facebook page.

DigitAll

Yes. I can help you. Write me on Facebook

DigitAll

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 :)

Kiriko

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

Max Mira

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

thor42

with lora or with Swarm UI?

Max Mira

i already did that and everything works fine

Shubz World

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?

Max Mira

ok will make flux lora workflow

DigitAll

Can u upgrade workflows to Flux and Flux+Lora?

Max Mira

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

Raymond Brettschneider

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.

DigitAll

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 ?

Claude GIROLET

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.

DigitAll

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.

Raymond Brettschneider

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 :)

Judi Godvliet

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.

Claude GIROLET

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.

DigitAll

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.

Claude GIROLET

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

DigitAll

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

Judi Godvliet

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

DigitAll

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

Judi Godvliet

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

21machines

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'

Yiluwidi Dreaming

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?

SoFiB34tz

No problem.

DigitAll

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??

Yiluwidi Dreaming

can you write me on facebook to help you?

DigitAll

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

SoFiB34tz

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

DigitAll

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

SoFiB34tz

Did you install the right workload for Visual Studio?

DigitAll

I had Visual Studio installed already before.

SoFiB34tz

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.

DigitAll

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?

SoFiB34tz

Can you write me on Facebook?

DigitAll

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.

Yiluwidi Dreaming

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.

DigitAll

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 . . .

Yiluwidi Dreaming

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]).

Shealan Forshaw

Ok. Onnit.

Yiluwidi Dreaming

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

DigitAll

Fixed. Download available

DigitAll

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

Yiluwidi Dreaming

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

DigitAll

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.

MHI Ir

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?

DigitAll

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

Oviyan

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

DigitAll

ok

DigitAll

Can you add a layerdiffusion with batch

Oviyan

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

DigitAll

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

Shealan Forshaw

Wat do you have to send to send screenshots?

DigitAll

I dont have FB - give me the fix pls

Yiluwidi Dreaming

write me on facebook please. i will help

DigitAll

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

Yiluwidi Dreaming

Thanks it works perfect... ☺️☺️

Patrick Nielsen

I only have 56 workflows

Andre Prior

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

DigitAll

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?

Chris "Broadway" Romero

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

DigitAll

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

Stacy Burnette

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

DigitAll

I cant seem to find the 1845 workflow after updating?

Patrick Nielsen

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

DigitAll

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

Stacy Burnette

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

DigitAll

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?

guni

Sorted it with the install of visual studio, thanks.

Richard Mathews

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

DigitAll

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

Richard Mathews

did it worked before update?

DigitAll

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.

Richard Mathews

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

DigitAll

Is there also a version for Mac?

Alex Fetter

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.

Elwaves

Thank you

AI_Guru

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.

thor42

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

AI_Guru

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.

DigitAll

Where should i find input prompts

AI_Guru

Where should i paste the style.csv in comfyui

AI_Guru

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

DigitAll

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

moti beny

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

DigitAll

Just double click on RUN VENV

DigitAll

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.

Elegost

Only one

DigitAll

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

Value Store Zone

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

Elegost

winrar app. 5 minute max

DigitAll

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

Elegost

supported little much as i could,

Shubz World

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

Zakaria Nada

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

Judi Godvliet

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

O.M.3

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

Tobe2d

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

DigitAll

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

Tobe2d

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

DigitAll

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?

Tobe2d

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 :(

Zom Boy

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

DigitAll

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

Richard Mathews

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)

Mikhail Ershov

Awesome, thanks for all the work!

Judi Godvliet

Yes, ControlNet workflows will be available soon.

DigitAll

Any plans to add img2img workflows?

Gabriel