| # LTX-2 Image-to-Video Adapter LoRA | |
| A high-rank LoRA adapter for [LTX-Video 2](https://github.com/Lightricks/LTX-Video) that substantially improves image-to-video generation quality. No complex workflows, no image preprocessing, no compression tricks -- just a direct image embedding pipeline that works. | |
| ## What This Is | |
| Out of the box, getting LTX-2 to reliably infer motion from a single image requires heavy workflow engineering -- ControlNet stacking, image preprocessing, latent manipulation, and careful node routing. The purpose of this LoRA is to eliminate that complexity entirely. It teaches the model to produce solid image-to-video results from a straightforward image embedding, no elaborate pipelines needed. | |
| Trained on **30,000 generated videos** spanning a wide range of subjects, styles, and motion types, the result is a highly generalized adapter that strengthens LTX-2's image-to-video capabilities without any of the typical workflow overhead. | |
| ### Key Specs | |
| | Parameter | Value | | |
| |-----------|-------| | |
| | **Base Model** | LTX-Video 2 | | |
| | **LoRA Rank** | 256 | | |
| | **Training Set** | ~30,000 generated videos | | |
| | **Training Scope** | Visual only (no explicit audio training) | | |
| ## What It Does | |
| - **Improved image fidelity** -- the generated video maintains stronger adherence to the source image with less drift or distortion across frames. | |
| - **Better motion coherence** -- subjects move more naturally and consistently throughout the clip. | |
| - **Broader generalization** -- performs well across diverse subjects and scenes without needing per-category tuning. | |
| ### A Note on Audio | |
| Audio was **not** explicitly trained into this LoRA. However, due to the nature of how LTX-2 handles its latent space, there are subtle shifts in audio output compared to the base model. This is a side effect of the training process, not an intentional feature. | |
| ## Usage (ComfyUI) | |
| 1. Place the LoRA file in your `ComfyUI/models/loras/` directory. | |
| 2. Add an **LTX-2** model loader node and load the base LTX-2 checkpoint. | |
| 3. Add a **Load LoRA** node and select this adapter. | |
| 4. Connect an **image embedding** node with your source image. | |
| 5. Add your text prompt and generate. | |
| ### Workflow | |
|  | |
| ## Examples | |
| Reference videos demonstrating the adapter's output quality: | |
| <video src="https://huggingface.co/MachineDelusions/LTX-2_Image2Video_Adapter_LoRa/resolve/main/assets/AnimateDiff_00740.mp4" autoplay loop muted playsinline></video> | |
| <video src="https://huggingface.co/MachineDelusions/LTX-2_Image2Video_Adapter_LoRa/resolve/main/assets/AnimateDiff_00774.mp4" autoplay loop muted playsinline></video> | |
| <video src="https://huggingface.co/MachineDelusions/LTX-2_Image2Video_Adapter_LoRa/resolve/main/assets/AnimateDiff_00777.mp4" autoplay loop muted playsinline></video> | |
| <video src="https://huggingface.co/MachineDelusions/LTX-2_Image2Video_Adapter_LoRa/resolve/main/assets/AnimateDiff_00778.mp4" autoplay loop muted playsinline></video> | |
| ## Model Details | |
| - **Architecture:** LoRA (Low-Rank Adaptation) applied to LTX-Video 2's transformer layers | |
| - **Rank 256** provides a high-capacity adaptation while remaining efficient to load and merge | |
| - **Training data** was intentionally diverse to avoid overfitting to any single domain, producing a general-purpose image-to-video adapter rather than a style-specific fine-tune | |
| ## License | |
| Please refer to the [LTX-Video license](https://github.com/Lightricks/LTX-Video) for base model terms. | |