⚡ Quick Answer
Image-to-video (img2vid) takes a single still image and generates motion around it, producing a short clip that animates that photo. The starting image anchors what the subject looks like — your prompt then guides how it moves.
This is a popular way to bring a product photo, a portrait, or an illustration to life without needing to describe the entire subject in words, the way text-to-video requires.
Where You'll See It
A Load Image node feeds into an image-conditioning input on a video model's loader or sampler, often labeled something like start_image. This is the key difference from text-to-video, which has no image input anywhere in the graph.
Quick Example
Upload a single photo of a person and run it through an image-to-video workflow with the prompt "turns head and smiles." The output is a few seconds of that exact person performing that motion, generated entirely from the one still image.
Frequently Asked Questions
See It In Action
Ready to animate a photo?
Our LTX-2 guide covers the full image-to-video setup, start to finish.
Published: 2026-09-17 · Last updated: 2026-09-17
Join the discussion
Sign in to leave a comment or reply
No comments yet
Be the first to share your thoughts!
