Instructions to use RuneXX/LTX-2.3-Workflows with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- LTX.io
How to use RuneXX/LTX-2.3-Workflows with LTX.io:
# Install the LTX-2 pipelines git clone https://github.com/Lightricks/LTX-2.git cd LTX-2 uv sync --frozen
# Download the weights from this repo, plus the Gemma text encoder hf download RuneXX/LTX-2.3-Workflows --local-dir models/LTX-2.3-Workflows hf download google/gemma-3-12b-it-qat-q4_0-unquantized --local-dir models/gemma-3-12b
# Fast pipeline (distilled model, no distilled LoRA needed) uv run python -m ltx_pipelines.distilled \ --distilled-checkpoint-path models/LTX-2.3-Workflows/<distilled-checkpoint>.safetensors \ --spatial-upsampler-path models/LTX-2.3-Workflows/<spatial-upsampler>.safetensors \ --gemma-root models/gemma-3-12b \ --prompt "A beautiful sunset over the ocean" \ --output-path output.mp4 # For image-to-video, add: --image path/to/image.jpg 0 0.8# HQ pipeline (two-stage, higher quality) uv run python -m ltx_pipelines.ti2vid_two_stages_hq \ --checkpoint-path models/LTX-2.3-Workflows/<checkpoint>.safetensors \ --distilled-lora models/LTX-2.3-Workflows/<distilled-lora>.safetensors 0.8 \ --spatial-upsampler-path models/LTX-2.3-Workflows/<spatial-upsampler>.safetensors \ --gemma-root models/gemma-3-12b \ --prompt "A beautiful sunset over the ocean" \ --output-path output.mp4 # For image-to-video, add: --image path/to/image.jpg 0 0.8 - Notebooks
- Google Colab
- Kaggle
Will you make workflows for Minimax H3?
Your LTX workflows are really neat,If you can make these neat workflows for the new SOTA Minimax H3 that would be great.
might ; - ) still testing it out. If it needs some alternative workflows for this and that usage.
might ; - ) still testing it out. If it needs some alternative workflows for this and that usage.
love your work man ❣️
might ; - ) still testing it out. If it needs some alternative workflows for this and that usage.
excited Best LTX workflows i used!
I'd love to see specialized H3 workflows like you did for LTX. The model is capable of so much, especially ref2v.
- Extend Video: focused workflow where you drop in a reference video with optional additional references
- Combine Video: feed it two clips and have it seamlessly join them, like VACE clip joiner
- Repair Video: give it a timestamp range and it seamlessly regenerates the video there (i.e. eliminates seams from joined videos)
- Repair Audio: have it regenerate the audio for a given video, or add audio to a silent video
- Fill In The Gap: provide two reference clips and have it generate a whole prompted section in between them
That's just off the top of my head. I can think dozens more. And yeah, I'm aware some of this can already be accomplished with the default Comfy workflows, but they're a bit tedious and rough around the edges. I personally like having separate workflows for separate jobs, like utility workflows.
I'm also aware that some of this isn't distinctly in the model's base capabilities, but people are finding clever ways around those limitations to accomplish these things. I'm sure you're already working on something awesome but I just thought I'd throw up some suggestions.