Скачать или смотреть ComfyUI LTX2.0 Uncensored Capability Enhancements（T2V, I2V and FLF2V）

#aiimages #comfy #comfyui #aivideo #runninghub #ltx
Join the Discord for Updates and Sharing
Discord:   / discord  
If you think the video is helpful, please like, share the video, and subscribe for more. I will update with more interesting things. Thank you.

ComfyUI LTX 2.0: Uncensored Mastery (T2V, I2V, & First-to-Last Frame Video)
Unlock the full, unrestricted potential of LTX 2.0. While the base model comes with default limitations, its open-source nature allows the community to push boundaries. In this guide, Faboro breaks down how to use uncensored merges and enhancement LoRAs to generate high-fidelity video with native synchronized audio—no mosaic, no limits.

We explore advanced workflows including Text-to-Video (T2V), Image-to-Video (I2V), and the powerful First and Last Frame (FLF2V) transition technique for ultimate cinematic control.

🚀 Key Features & Workflow Highlights:
LTX2-Rapid-Merges: We use Phr00t’s uncensored model base, which integrates NSFW-capable LoRAs directly into the weights for better anatomy and concept adherence.

Anatomy Enhancement LoRAs: Learn how to stack specific LoRAs to fix anatomical details and refine the "forbidden" parts that base models often struggle with.

First and Last Frame (FLF2V): A game-changing feature! Upload a starting image and an ending image, and LTX 2.0 will generate a smooth, consistent transition between them.

The VAE Fix: Important! If your videos look grainy or distorted, you might have the "wrong" official VAE. We show you how to identify and install the correct 24kHz version.

Cloud Safety (Runninghub Trick): Generate uncensored content on the cloud without getting banned. Use the Super-SecureMediaProtection node to encode your results into a "duck image" and decode them locally on your PC.

Prompt Enhancement: Don't get stuck with complex syntax. Use our built-in AI prompt refiner to expand simple ideas into detailed, LTX-optimized descriptions.

🛠️ Resources & Download Links:
All the model details are within the workflow.
Workflow(Download or run on cloud computer Runninghub.ai, 1000 free credits first time, and 100 daily login credits with 4090 48GB)
LTXV2.0 Uncensored Video Generation Workflow(Online Version)
https://www.runninghub.ai/post/201381...
LTXV2.0 Uncensored Video Generation Workflow(Local Version)
https://civitai.com/models/2329764?mo...

LTX2 prompting guide:
https://ltx.io/model/model-blog/promp...

Phr00t model:
https://huggingface.co/Phr00t/LTX2-Ra...
KJ model:
https://huggingface.co/Kijai/LTXV2_comfy

NSFW Decoding Tool: [Link to Duck Image Decoder]
https://github.com/copyangle/SS_tools
http://duck.airush.top:81/

Custom Nodes:
https://github.com/Lightricks/ComfyUI...
https://github.com/kijai/ComfyUI-KJNodes
https://github.com/Kosinkadink/ComfyU...
https://github.com/filliptm/ComfyUI_F...
https://github.com/Windecay/ComfyUI-R...

ComfyUI Installation Guide:
   • Faster ComfyUI With Sage Attention and Tri...  
https://comfyui-wiki.com/en/install/i...

Video Timeline:
00:00 - Introduction and Latest Update
02:48 - ComfyUI and Model Installation Guide(Newbie only)
03:30 - Workflow Explanation
11:08 - Test Examples
13:49 - Final Words