language:
- en
- de
- es
- fr
- ja
- ko
- zh
- it
- pt
license: other
license_name: ltx-2-community-license-agreement
license_link: https://github.com/Lightricks/LTX-2/blob/main/LICENSE.md
pipeline_tag: image-to-video
arxiv: 2601.03233
tags:
- image-to-video
- text-to-video
- video-to-video
- image-text-to-video
- audio-to-video
- text-to-audio
- video-to-audio
- audio-to-audio
- text-to-audio-video
- image-to-audio-video
- image-text-to-audio-video
- multimodal
- ltx-video
- lightricks
- diffusion-single-file
pinned: true
extra_gated_description: >-
By clicking "Agree and Access" you acknowledge the [Privacy
Policy](https://static.lightricks.com/legal/Privacy%20Policy%20-%20LTX%20Platform.pdf)
and consent to receive offers and updates including targeted and personalized
advertisements. You can unsubscribe at any time.
extra_gated_button_content: Agree and Access
Commercial and production use at no cost under the LTX-2.x Community License. Transfer of fine-tunes may require a paid license, in accordance with the LTX-2.x Community License.
Read the DocumentationPaid Commercial Use Agreement for LTX-2.x with full weights, engineering support, LoRAs, and flexible deployment options. To learn about all licensing options, talk to an expert.
Talk to a Commercial Licensing ExpertCommercial and production use at no cost under the LTX-2.x Community License. Transfer of fine-tunes may require a paid license, in accordance with the LTX-2.x Community License.
Read the DocumentationPaid Commercial Use Agreement for LTX-2.x with full weights, engineering support, LoRAs, and flexible deployment options. To learn about all licensing options, talk to an expert.
Talk to a Commercial Licensing ExpertLTX-2.5 Pre-Trained is the base checkpoint produced during pre-training, before supervised fine-tuning and other post-training used for the main LTX-2.5 release. It is intended for researchers and developers creating specialized models and can be fine-tuned using the standard LTX Trainer. For general-purpose video generation, use the main LTX-2.5 checkpoint.
Model family & checkpoints
LTX-2.5 Pre-Trained consists of a single bundled checkpoint — the 22B DiT together with the video VAE, audio VAE, vocoder and duration head — plus the matching Gemma4 text encoder.
Transformers (DiT)
| File | Notes |
|---|---|
ltx-2.5-22b-pt-bf16.safetensors |
Pre-trained DiT (bf16), 43 GiB. Single-file bundle — also contains the video VAE, audio VAE, vocoder, duration head and text-embedding projections. |
Other components
| File | Notes |
|---|---|
ltx-2.5-22b-gemma4-12b/ |
Gemma4 12B text encoder (bf16), HF Transformers format. Required — the checkpoint expects gemma_version gemma4-12b-ltx-v1. |
| Video VAE, audio VAE, vocoder, duration head | Bundled inside the checkpoint above — no separate files. |
Training & fine-tuning
This is the pre-training checkpoint, before the supervised fine-tuning and post-training used for the main LTX-2.5 release. Fine-tune it with the LTX-2 Trainer, paired with the text encoder above.
Limitations
- This model is not intended or able to provide factual information.
- As a statistical model, this checkpoint may amplify existing societal biases.
- This checkpoint is not optimized in the same way as the main LTX-2.5 generation model. It does not include the supervised fine-tuning and subsequent post-training used for the main release.
- Fine-tuned derivatives may introduce additional capabilities, limitations and safety risks and should be evaluated independently.
Citation
@article{hacohen2025ltx2,
title={LTX-2: Efficient Joint Audio-Visual Foundation Model},
author={HaCohen, Yoav and Brazowski, Benny and Chiprut, Nisan and Bitterman, Yaki and Kvochko, Andrew and Berkowitz, Avishai and Shalem, Daniel and Lifschitz, Daphna and Moshe, Dudu and Porat, Eitan and Richardson, Eitan and Guy Shiran and Itay Chachy and Jonathan Chetboun and Michael Finkelson and Michael Kupchick and Nir Zabari and Nitzan Guetta and Noa Kotler and Ofir Bibi and Ori Gordon and Poriya Panet and Roi Benita and Shahar Armon and Victor Kulikov and Yaron Inger and Yonatan Shiftan and Zeev Melumian and Zeev Farbman},
journal={arXiv preprint arXiv:2601.03233},
year={2026}
}