This is the Image to Video workflow for Wan2.2.

We've added a new “End Image” feature to the workflow released last time.

Upload your Start Image and End Image, then enter any prompt to generate a video that ends with the frame from the End Image.

This feature allows you to prepare your desired images in advance, enabling video generation that reflects more detailed creative intent.

You can also create a looping video by making the last image the same as the first image.

The LoRA configuration is Lightx2v + FusionX. You can customize it to your preferences.

It's a simple yet versatile workflow with plenty of customization potential.

TIME COST

For Free Generation users, here are the estimated time costs.

When generated with the default settings (without specifying the final frame), generation took approximately 2 minutes 30 seconds to 3 minutes.

If you specify an final frame, the generation time will be extended, and may take 3 to 7 minutes.

Generation times may vary depending on the settings. Additionally, generation times may change during peak service periods. Please consider this time as an estimate only.

HOW TO USE

1. Upload an image.

If you don't need an end image, bypass it. (See below.)

2. Enter a prompt.

What kind of video do you want? Negative prompts are optional.

3. Video Length

Specify the number of frames. This is the number of images per FPS (frames per second). The default FPS is 16 and the length is 81 frames. This creates a video of approximately 5 seconds.

4. Resolution Settings

Set the long side of the resolution. Increasing the resolution beyond the default may result in generation failure or significantly longer generation times, so only tweak this setting if you have the knowledge to change other settings.

5. Loop Count

This function combines multiple loops of a video. In other words, setting this to 1 will create a single video by repeating the exact same 5-second video twice. This setting is rarely necessary.

6. Press the Generate button.

LoRA

You can disable unused nodes by right-clicking and selecting "Bypass." If you don't use an end image, i.e., if you want the AI to generate the final frame of the video, bypass this.

The base model and LoRA nodes are located at the top of the workflow.

Only FusionX and Lightx2 are enabled in the preset.

Purple nodes are bypassed, so if you want to use them, right-click and select "Bypass."

The maximum number of LoRA nodes that can be used at one time is 5.

Known issues.

1. Pipe Bug

Occasionally, a workflow may fail to load, causing pipes (wires connecting nodes that run freely across the screen) to become disconnected.

Reloading the workflow will fix this.

2. VRAM Bug

Occasionally, an error due to insufficient VRAM may occur.

This is an error on the SeaArt site's side, and it often occurs during busy times. Please try again later.

3. Model and LoRA Loading Bug

This occurs randomly. Regenerating the model may resolve the issue.

WAN2.2 Bild in Video

Bewertungen & Kommentare