ComfyUI-LTXVideoLoRA

A set of custom nodes enabling LoRA support for LTX Video in ComfyUI.

25.02.2025 ⭐ NEW ⭐

• Update the LTXV LoRA Loader node for the GGUF support as a generic way
• Remove useless LTXV Checkpoint Loader with LoRA node

Please refer to the quantized versions of :

• calcuis - GGUF quantized and fp8 scaled versions of LTX-Video (here)
• city96 - direct GGUF conversion of Lightricks/LTX-Video (here)

08.02.2025 - First release

• ~~Add LoRA support as a individual LTXV LoRA Loader node > for Lightricks ComfyUI-LTXVideo~~
• ~~Add LoRA support inside a LTXV Checkpoint Loader with LoRA node > for log(td) ComfyUI-LTXTricks~~
• Add LoRA selector node that can be chained using multiple LTXV LoRA Selector

Installation

Installation via ComfyUI-Manager

Installation via ComfyUI-Manager is preferred. Simply search for ComfyUI-LTXVideoLoRA in the list of nodes.

Manual installation

Simply clone this repository to custom-nodes folder in your ComfyUI installation directory.

Usage

LTXV LoRA with the ComfyUI-LTXVideo nodes

[!CAUTION] Because the LTXV Loader official node use a specific model structure, you need to put the ltxv_loader_as_input option as true.

LTXV LoRA with the ComfyUI-LTXTricks nodes

[!NOTE] Here the input node can be the 'Load Diffusion Model' node or the 'Load Checkpoint' node as well.

LTXV LoRA with the GGUF Loader nodes

[!TIP] As you can see, you can use the GGUF with the LoRA loader followed by the modified LTX model by log(td)'s LTXTricks nodes.

LoRA Training

For a blazing fast training of your LTXV LoRA you can use:

• a-r-r-o-w's finetrainers (here)

Credit and greetings

The main code is inspired by:

• comfyanonymous ComfyUI (here)
• Lightricks ComfyUI-LTXVideo (here)
• log(td) ComfyUI-LTXTricks (here)
• kijai ComfyUI-HunyuanVideoWrapper (here) for the LoRA Selector / Block Edit nodes.