Home Creators Posts Import Register Favorites Logout

Downloads

Content

Hello my friends,

LTX 2.3 has been out for a couple of days now, and I did everything I could to implement it into my installer as fast as possible so you can get your hands on it! I marked the "Old" LTX 2 as Legacy so you don't mix it up.

As usual, the Installer comes with all the models, custom nodes, and 4 workflows ready to go. The official YouTube video will take me a little more time as I am still recording, but for now: here you go, have some fun!

I also want to tease a couple of massive improvements coming to Diffusion Nexus. I don’t have a release date for this yet because there are still a lot of dangling cables under the hood. Honestly, I need a quick break from coding to touch grass (or just render some waifus). But here is what you can look forward to:

✨ A Whole New Module: The LoRA ViewerThis tool will scan your locally downloaded LoRAs, retrieve the data directly from Civitai, and store it locally. That means you get example images, descriptions, trigger words, and tags right at your fingertips.

It also lets you easily download missing versions! Let's say you have an ZImage Turbo version of a LoRA, and a Qwen version drops. With just 2 clicks, the software will download it for you.

🔍 Improved Image Viewer & Metadata ExtractionGot a cool image and want to reuse the exact prompt? No problem. My Image Viewer now reads the metadata directly from the image.Works with 100% of Automatic1111/Forge UI images and 90% of ComfyUI images (some workflows are just a bit too complicated for my algorithm to figure out).

🔄 ComfyUI Auto-UpdaterMy tool now checks if there is an update for ComfyUI, and if there is, it gets rolled out automatically.

🖥️ Start Your UIs from AnywhereYou can now open the new Unified Logger and use it to launch your UIs from anywhere.

✍️ Captioning Autocomplete & Spell CheckerSometimes you want to recaption some images or improve the writing. My new built-in spell checker and autocomplete will help you easily find and keep the right words.

Enjoy the LTX 2.3 update, and stay tuned for more!

Don’t forget to vote for future content (Multiple answers allowed) Post is open until Mar 14thhttps://www.patreon.com/posts/sneak-peek-of-i-151557474And if you are an AI Art Architect or higher you can vote twice!

https://www.patreon.com/posts/here-is-your-for-151558450

Best Regards

Chris

Files

Comments

Logovore

As a filthy casual, it would be nice if you included a tiny summary of whatever it is that you’re showcasing. Love your content!!

AI Knowledge Central

Hi, sorry for the confusion. I am showcasing my Diffusion Nexus software that started as a Lora Dataset manager or new features that will come in the next update. Since its only available for AI Art Collector's and up it may have been a better idea to put in a separate post. I try to avoid to spam my Patreon feed with two many previews and updates that not actually deliver content.

Wang Buffet

Can Diffusion Nexus run on Mac OS?

AI Knowledge Central

Hi for now its Windows only but Linux is in testing. I have no device for testing Mac OS.

Pixel Freak

I got your installer and everything seemed to work great. However, I get this error when trying to run a job. I get it on Your LTX 2 and 2.3 workflow. Am I doing something wrong or did comfy update something? I'm a noob. Unet Loader (GGUF) ValueError: cannot reshape array of size 3623914 into shape (4096,2304)

AI Knowledge Central

Hi Mike, thanks for your support. The error you are seeing is either a model mismatch meaning somewhere in the workflow a incompatible model is selected or your ComfyUI is not up to date. Can you write me a DM please with the full output of your terminal window from the moment you start up ComfyUI until you run into the error?

Daniel Bostic

Hey, i got the DiffusionNexus.Installer, im not sure what im doing wrong but i can't seem to get it to work :( i had to reinstall my confy ui since my last install corrupted and i didn't want to go through the tedious installation process again, but when i tried the new Comfy install i ran into a few issues: 1. it wasnt able to clone the repo into my folder; it said that there was no git repo.

Jérémy de Benedictis

Is there a possibility to add a controlNet to your image to video ltx NSFW workflow? I can't get it to work :'( I'd like to use openPose

Stevie

Hi, probably missed the details but can this run offline?

AI Knowledge Central

You need to be online for the installation because it will download quite some GB of Models. but after that everything runs offline.

Flap Jack

flux220kleinE58AA8E6BCABE8BDACE5.O9j8.safetensors where can i find this