Instructions to use Lightricks/LTX-2.3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use Lightricks/LTX-2.3 with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline from diffusers.utils import load_image, export_to_video # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("Lightricks/LTX-2.3", dtype=torch.bfloat16, device_map="cuda") pipe.to("cuda") prompt = "A man with short gray hair plays a red electric guitar." image = load_image( "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/guitar-man.png" ) output = pipe(image=image, prompt=prompt).frames[0] export_to_video(output, "output.mp4") - LTX.io
How to use Lightricks/LTX-2.3 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 Lightricks/LTX-2.3 --local-dir models/LTX-2.3 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/<distilled-checkpoint>.safetensors \ --spatial-upsampler-path models/LTX-2.3/<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/<checkpoint>.safetensors \ --distilled-lora models/LTX-2.3/<distilled-lora>.safetensors 0.8 \ --spatial-upsampler-path models/LTX-2.3/<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
Clarification on LTX-2 Community License applicability to a public university
Hello,
I am affiliated with a public university in Japan and am evaluating LTX-2.3
for a research project that would run entirely locally (no cloud deployment,
no redistribution of outputs).
I have read the LTX-2 Community License Agreement and would like to confirm
one point before proceeding. Section 2 states that "Entities with annual
revenues of at least $10,000,000" are required to obtain a paid commercial
use license. The license text does not explicitly define whether this
revenue threshold is intended to apply to non-profit public educational
institutions conducting non-commercial academic research, as distinct from
commercial entities.
Could you clarify:
- Does the $10,000,000 annual revenue threshold apply to a public
university's overall institutional budget, or is it intended only for
commercial/revenue-generating entities? - Is non-commercial academic research use by a university covered under
the free license grant in Section 2, regardless of the institution's
total budget?
I would appreciate any written confirmation or pointer to existing guidance
on this question, as it will determine whether we can proceed with LTX-2.3
for research purposes.
Thank you for your time.
Best regards,
HS. Choi
Dear HS. Choi,
You do not need a paid commercial use agreement for the use you describe.
Non-commercial academic research by your university, run entirely locally with no redistribution of outputs and no revenue arising from the use, falls within the free license grant regardless of the institution's overall budget. You may proceed with LTX-2.3 on that basis.
Please keep in mind that this free-use permission is limited to genuinely non-commercial purposes. A paid license would be required if the use later moves into a production environment, generates direct or indirect revenue, involves direct interaction with end users, or is used to train or fine-tune a model for commercial use.
Best regards,
LTX Team
Thank you for your response.
One follow-up, if I may: in our planned study, research participants directly operate the local system themselves — they type prompts and generate videos on the machine in real time, as part of a supervised academic experiment (no public access, no product, no revenue).
Would this participant-facing interaction with the local model count as "direct interaction with end users" under the license, or does that term refer specifically to deploying LTX-2.3 within a public-facing product or service?
Thank you again for your help.
If used within a planned study, by research participants only, as part of a supervised academic experiment (no public access, no product, no revenue). Our previous response still applies.
Best regards,
LTX Team