README.md
5.3 KB · 126 lines · markdown Raw
1 ---
2 language:
3 - en
4 - de
5 - es
6 - fr
7 - ja
8 - ko
9 - zh
10 - it
11 - pt
12 library_name: diffusers
13 license: other
14 license_name: ltx-2-community-license-agreement
15 license_link: https://github.com/Lightricks/LTX-2/blob/main/LICENSE
16 pipeline_tag: image-to-video
17 arxiv: 2601.03233
18 tags:
19 - image-to-video
20 - text-to-video
21 - video-to-video
22 - image-text-to-video
23 - audio-to-video
24 - text-to-audio
25 - video-to-audio
26 - audio-to-audio
27 - text-to-audio-video
28 - image-to-audio-video
29 - image-text-to-audio-video
30 - ltx-2
31 - ltx-2-3
32 - ltx-video
33 - ltxv
34 - lightricks
35 pinned: true
36 demo: https://console.ltx.video/playground/
37 ---
38
39 # LTX-2.3 FP8 Model Card
40
41 **This is the FP8 versions of the LTX-2.3 model. All information below is derived from the base model.**
42
43 This model card focuses on the LTX-2.3 model, which is a significant update to the [LTX-2 model](https://huggingface.co/Lightricks/LTX-2) with improved audio and visual quality as well as enhanced prompt adherence.
44 LTX-2 was presented in the paper [LTX-2: Efficient Joint Audio-Visual Foundation Model](https://huggingface.co/papers/2601.03233).
45
46 💻💻 **If you want to dive in right to the code - it is available [here](https://github.com/Lightricks/LTX-2).** 💾💾
47
48 LTX-2.3 is a DiT-based audio-video foundation model designed to generate synchronized video and audio within a single model. It brings together the core building blocks of modern video generation, with open weights and a focus on practical, local execution.
49
50 [![LTX-2.3 Open Source](https://img.youtube.com/vi/o-7us-BR_gQ/maxresdefault.jpg)](https://youtu.be/o-7us-BR_gQ)
51
52 # Model Checkpoints
53
54 | Name | Notes |
55 |------------------------------------|--------------------------------------------------------------------------------------------------------------------|
56 | ltx-2.3-22b-dev-fp8 | The full model, flexible and trainable, in fp8 |
57 | ltx-2.3-22b-distilled-fp8 | The distilled version of the full model, 8 steps, CFG=1, in fp8 |
58
59 ## Model Details
60 - **Developed by:** Lightricks
61 - **Model type:** Diffusion-based audio-video foundation model
62 - **Language(s):** English
63
64 # Online demo
65 LTX-2.3 is accessible right away via the [API Playground](https://console.ltx.video/playground/).
66
67 # Run locally
68
69 ## Direct use license
70 You can use the models - full, distilled, upscalers and any derivatives of the models - for purposes under the [license](./LICENSE).
71
72 ## ComfyUI
73 We recommend you use the built-in LTXVideo nodes that can be found in the ComfyUI Manager.
74 For manual installation information, please refer to our [documentation site](https://docs.ltx.video/open-source-model/integration-tools/comfy-ui).
75
76 ## PyTorch codebase
77
78 The [LTX-2 codebase](https://github.com/Lightricks/LTX-2) is a monorepo with several packages. From model definition in 'ltx-core' to pipelines in 'ltx-pipelines' and training capabilities in 'ltx-trainer'.
79 The codebase was tested with Python >=3.12, CUDA version >12.7, and supports PyTorch ~= 2.7.
80
81 ### Installation
82
83 ```bash
84 git clone https://github.com/Lightricks/LTX-2.git
85 cd LTX-2
86
87 # From the repository root
88 uv sync
89 source .venv/bin/activate
90 ```
91
92 ### Inference
93
94 To use our model, please follow the instructions in our [ltx-pipelines](https://github.com/Lightricks/LTX-2/blob/main/packages/ltx-pipelines/README.md) package.
95
96 ## Diffusers 🧨
97
98 LTX-2.3 support in the [Diffusers Python library](https://huggingface.co/docs/diffusers/main/en/index) is coming soon!
99
100 ## General tips:
101 * Width & height settings must be divisible by 32. Frame count must be divisible by 8 + 1.
102 * In case the resolution or number of frames are not divisible by 32 or 8 + 1, the input should be padded with -1 and then cropped to the desired resolution and number of frames.
103 * For tips on writing effective prompts, please visit our [Prompting guide](https://ltx.video/blog/how-to-prompt-for-ltx-2)
104
105 ### Limitations
106 - This model is not intended or able to provide factual information.
107 - As a statistical model this checkpoint might amplify existing societal biases.
108 - The model may fail to generate videos that matches the prompts perfectly.
109 - Prompt following is heavily influenced by the prompting-style.
110 - The model may generate content that is inappropriate or offensive.
111 - When generating audio without speech, the audio may be of lower quality.
112
113 # Train the model
114
115 Currently it is recommended to train the bf16 model. Recipes for training the fp8 model are welcome as community contributions.
116
117 ## Citation
118
119 ```bibtex
120 @article{hacohen2025ltx2,
121 title={LTX-2: Efficient Joint Audio-Visual Foundation Model},
122 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},
123 journal={arXiv preprint arXiv:2601.03233},
124 year={2025}
125 }
126 ```