Home Creators Posts Import Register Favorites Logout

Downloads

Content

Update 28th Dec Qwen Image Edit and more.Watch the Video:https://youtu.be/Ic3oMRCYFDo

Update 21th Dec 2025-Added FlashVSR Workload + minor bug fixes.

Update 20th Dec 2025Ver 0.21 Is up. I included some bug fixes and also some minor features:

By User Request I implemented a feature that lets you store downloaded models in a different location. I also like to story my models in a decentralized environment. That avoids redownloading everything every time. In theory my software should also skip duplicate downloads. (can be buggy since its a new feature).

Rename the shortcut to a name of your choice.

You can now filter by Workload Type. This will become useful in the future when I include A1111,Forge, AI-Tookit and more, also the Workload shows for what Software and what version. The Version thing is more for me.

Original Post

Hello my amazing Patrons,

I’ve been quiet for the last couple of days. For you, that usually means good things are happening. It means I’ve been hyper-focused deep in my workshop crafting something special. I realized I never made a proper Wan 2.2 One-click installer. Well here it is and more!

What I brought to you today, you won't find on any other Gen-AI Patreon!

I have always been dissatisfied with standard "one-click installers." Why? Because they are brittle, user-unfriendly, ugly, and often cause more pain than they solve. So, I said "fuck it." Not many of you know this, but I’m a software developer in my day job—not just a "vibe coder," but the real deal.

So, I decided to code a solution to make your life easier.

Introducing: The Workload InstallerThis is one piece of software to replace all my brittle batch files. Currently, it integrates the VAN 2.2 workflow, but this is just the beginning.

🎁 The Holiday GiftTo celebrate what an amazing year this has been, I’m giving this installer away for FREE.

Download it before January 3rd.

After that, it will become exclusive to the AI Apprentice Tier and up.

What else is cooking?

The big Z Image Turbo LoRA training video (Patreon-exclusive cut) is being edited right now

Discord server is finally happening – rolling out to higher tiers first, then gradually to everyone (be gentle, it’s my first one)

The LoRA training workloads will of course install through the new tool – no batch-file nightmares ever again

A couple of surprises I’m keeping secret for now
 because I like watching you guess in the comments 😏

That’s it from the workshop for today.Thank you again for making all of this possible. Seriously. You’re legends.

Chris

Files

Next Level Easy-installer!

This is how you Install ComfyUI, A1111, Forge, AI-Toolkit and many more nowadays. Download it here: https://www.patreon.com/posts/huge-thank-you-2-146207704?utm_medium=clipboard_copy&utm_source=copyLink&utm_campaign=postshare_creator&utm_content=join_link

Comments

newtoai

Do you have also wf for wan 2.1 with wan 2.1 loras support? Or wan 2.2 wf that support both 2.1 and 2.2 loras? Thanks!

AI Knowledge Central

This is a great Idea, I will put wan 2.1 on my list of things to put in there. But tbh my current focus is video Upscale and shortly after lora training

newtoai

Thanks! Is there any lora loader node that you could add in your workflow that can pre load and list all the trigger words of the loras we upload?

puk

Super Arbeit, fĂŒge bitte noch 96GB RAM fĂŒr RTX Pro 6000 hinzu.

James Charleston II

This is very cool. Just a suggestion that may help in the future would be to have option of central model storage. This way, your downloader could detect if the models are already downloaded and they are not duplicated across on new builds. Outside of that, this looks amazon.

AI Knowledge Central

Thanks for the Feedback. This is a great Idea! consider it implemented in the next update. Probably under the advanced section. for starters i will make it a simple name check. In the future i will implement a hash comparison.

AI Knowledge Central

FĂŒr diesen speziellen workload nutze ich die GGUF Quants. Da ist Q8 also 32GB leider das höchste der GefĂŒhle. Ich werde aber sehen was ich machen kann fĂŒr zukĂŒnftige Workloads.