Animating Images with LTX-2 Image-to-Video

Lip-Sync & Talking Character Example

Artistic Animation Example

Initial Configuration for LTX-2 Image-to-Video Workflow

1. Select 'Distilled Workflow' for more optimized generation speeds.
2. Go to the first card (Load Models) and click the dropdown menu for each model.
3. Select the downloaded model files for each respective section.
4. In the 'Gemma 3' section, choose the '4-bit model one' version.
5. Under 'LTXV path', ensure you select the matching 'Distilled' model in both provided slots.

Adjusting Video Parameters and Image Inputs

1. Locate the 'Length' setting to define the video duration in frames.
2. Set the frame value based on your desired duration (e.g., 121 frames for ~5 seconds, or 200 frames for a longer duration).
3. Click on the 'Upload Image' section to upload the static image you want to use as the starting frame.
4. Ensure the uploaded image has a clear subject to make animation easier (such as a person speaking or singing).

Entering Prompts and Generating Video

1. Enter the action description or dialogue into the text prompt box.
2. To speed up the generation process, 'Bypass' or close the 'Enhancer' node if necessary.
3. Press the 'Run' button to start the rendering process.
4. Wait until the video appears in the preview window.
5. To add camera movement effects, activate 'LoRA' (such as Dolly Left) and connect it to 'Stage 1' (low-res) and 'Stage 2' (upscale).