Instructions to use Lightricks/LTX-2.5 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusion Single File
How to use Lightricks/LTX-2.5 with Diffusion Single File:
# No code snippets available yet for this library. # To use this model, check the repository files and the library's documentation. # Want to help? PRs adding snippets are welcome at: # https://github.com/huggingface/huggingface.js
- LTX-2
How to use Lightricks/LTX-2.5 with LTX-2:
# Install the LTX-2 pipelines git clone https://github.com/Lightricks/LTX-2.git cd LTX-2 uv sync --extra natten
# Download weights from this repo # Substitute filenames from this repo's "Files and versions" if they differ hf download Lightricks/LTX-2.5 \ diffusion_models/<distilled-transformer>.safetensors \ text_encoders/gemma4-12b-with-proj-ltx-2.5-bf16.safetensors \ vae/<video-vae>.safetensors \ vae/<audio-vae>.safetensors \ latent_upscale_models/<spatial-upsampler>.safetensors \ latent_upscale_models/<temporal-upsampler>.safetensors \ --local-dir models/LTX-2.5 # DFR requires the detailing IC-LoRA (separate repo; strength is fixed at 0.5) hf download Lightricks/LTX-2.5-22b-IC-LoRA-Pixel-Spatial-Upscaler --local-dir models/LTX-2.5-22b-IC-LoRA-Pixel-Spatial-Upscaler# Distilled LTX-2.5 pipeline (fast) uv run python -m ltx_pipelines.distilled \ --transformer-path models/LTX-2.5/diffusion_models/<distilled-transformer>.safetensors \ --text-encoder-path models/LTX-2.5/text_encoders/gemma4-12b-with-proj-ltx-2.5-bf16.safetensors \ --video-vae-path models/LTX-2.5/vae/<video-vae>.safetensors \ --audio-vae-path models/LTX-2.5/vae/<audio-vae>.safetensors \ --spatial-upsampler-path models/LTX-2.5/latent_upscale_models/<spatial-upsampler>.safetensors \ --num-frames 121 \ --prompt "A beautiful sunset over the ocean" \ --output-path output.mp4 # For image-to-video, add: --image path/to/image.jpg 0 0.8# DFR pipeline (higher detail fidelity; optional temporal 2x/4x) uv run python -m ltx_pipelines.dfr_pipeline \ --transformer-path models/LTX-2.5/diffusion_models/<distilled-transformer>.safetensors \ --text-encoder-path models/LTX-2.5/text_encoders/gemma4-12b-with-proj-ltx-2.5-bf16.safetensors \ --video-vae-path models/LTX-2.5/vae/<video-vae>.safetensors \ --audio-vae-path models/LTX-2.5/vae/<audio-vae>.safetensors \ --spatial-upsampler-path models/LTX-2.5/latent_upscale_models/<spatial-upsampler>.safetensors \ --temporal-upsampler-path models/LTX-2.5/latent_upscale_models/<temporal-upsampler>.safetensors \ --detailing-lora models/LTX-2.5-22b-IC-LoRA-Pixel-Spatial-Upscaler/ltx-2.5-22b-ic-lora-pixel-spatial-upscaler-x2-1.0.safetensors \ --spatial-upscalings 1 \ --temporal-upscalings 1 \ --height 1088 \ --width 1920 \ --num-frames 121 \ --prompt "A beautiful sunset over the ocean" \ --output-path output.mp4 # For 4K: --spatial-upscalings 2 --width 3840 --height 2176 # For image-to-video, add: --image path/to/image.jpg 0 0.8 - Notebooks
- Google Colab
- Kaggle
Please tell the settings for full model not just distilled.
I know distilled is 8 step model but what if I want to run the full model? how many steps? cfg? anything? For comfyui.
I would guess full model requires 20-30 steps to get better quality.
Testing it soon!
Yes please because I ran till 50 steps and still getting bad outputs I mean just bad.
Yes please because I ran till 50 steps and still getting bad outputs I mean just bad.
I recommend to use sampler res2s both stages or res2s 1st and lcm 2nd
I know distilled is 8 step model but what if I want to run the full model? how many steps? cfg? anything? For comfyui.
My workflow for dev int is embedded!
I just tried LTX 2.5 FULL/DEV 20-steps (multi-shot, Ref2Vid), and it was one of the worst video outputs I've seen in a while. It wasn't glitchy or ignored the prompt. It just looked cheap and plastic--like an early version of AI video from two years ago. And it took me 43 minutes on an RTX 3090/64gb RAM. I had high hopes, but was majorly disappointed.
I just tried LTX 2.5 FULL/DEV 20-steps (multi-shot, Ref2Vid), and it was one of the worst video outputs I've seen in a while. It wasn't glitchy or ignored the prompt. It just looked cheap and plastic--like an early version of AI video from two years ago. And it took me 43 minutes on an RTX 3090/64gb RAM. I had high hopes, but was majorly disappointed.
In what workflow? The supplied ones are built for distilled with the sigma's and all the stuff i think
My workflow for dev int is embedded!
your dev workflow is not correct I'm afraid. there are several issues with the setup down the line but those don't matter because first and most importantly you've used the int8 21Gb nerfed model, not the full dev one bf16 the one that is 42Gb.
unfortunately I've not been able to find a correct template for the full power model (it will require an RTX Pro 6000 with 96 GB VRAM) so if anyone mentions an 3090 or 4090 then you know the setup is not ok. could work on a 5090 for shorter videos but I need to test that once we figure out how this full dev comfyui workflow works