How to Create AI Influencer Videos

AI video is what brings an AI influencer to life. While photos show your character in static moments, videos show them moving, gesturing, and expressing — the kind of content that performs well on social media.

Larpable generates videos using motion control technology: you provide an image of your AI identity and a reference video that defines the motion. Larpable then animates your identity to match.

How Video Generation Works

Larpable's video generation uses a two-input approach:

1. 1Image input — a photo of your AI identity (either a reference photo or a generated image). This defines what the character looks like.
2. 2Reference video — a short video showing the motion you want. This defines how the character moves.

The system takes the visual appearance from the image and the motion pattern from the reference video, then combines them into a new video where your AI identity performs the motion from the reference.

Step 1: Prepare Your Image

Start with a clear image of your AI identity. You can use a generated photo or one of your original reference photos. The image should show your identity clearly, preferably in a pose that's somewhat similar to the starting pose of your reference video.

Step 2: Choose a Reference Video

The reference video defines the motion. It should be a short clip (5–10 seconds works well) showing the movement you want your AI identity to perform. This could be:

• A person turning their head or looking around
• A person walking or posing
• A person making hand gestures
• A person smiling or changing expression

Step 3: Generate the Video

On the canvas workspace, connect your image and your reference video to a Video Generation node. You can add an optional prompt to guide the result. Click generate, and Larpable will process the video.

Video generation takes longer than image generation — typically several minutes. You can continue working while the video processes, and it will appear in your library when ready.

Best Practices for Realistic Videos

• Use reference videos with smooth, natural motion — jerky or fast movements produce less realistic results
• Keep reference videos short (5–10 seconds) for best results
• Match the image's framing to the reference video's framing — if the reference video is a close-up, use a close-up image
• Choose reference videos with good lighting similar to your image
• Avoid reference videos with complex occlusions (objects passing in front of the person)

Common Issues

The motion looks unnatural

Try a different reference video with smoother, simpler motion. Complex or fast movements are harder for the system to reproduce accurately.

The face doesn't match my identity

Make sure you're using a clear, front-facing image of your identity as the image input. If the image is at an unusual angle, the result may not match well.

Next Steps

Once you're creating videos, learn how to grow your AI influencer and create social media content that engages your audience.