Image to video models and APIs.
Animate a still: a product photo, a portrait or a first frame becomes a video clip. Includes reference-to-video models and ready-made video effects.
How to pick an image to video model
Image to video starts from a still you already have, so the look is fixed and the model only adds motion. It is the most predictable way to get on-brand video.
Animate one image
Seedance 2.5 and Kling V3 both start from your frame, and Seedance can also take a last frame, so you decide where the shot ends. PixVerse C1 animates a still into a clip of up to 15 seconds at 1080p, with optional audio.
Keep a character or product the same
Reference-to-video models take several images, sometimes clips and audio too, and keep the subject consistent through the shot. Try Wan 3.0 R2V or Seedance 2.5 Reference-to-Video.
Copy motion from a video
Kling V2.6 Motion Control takes your image and a reference video, and moves your character the way the reference moves.
Made for e-commerce
Product Ads turns a product photo into an animated ad with more than 100 presets. Product Model Video Generator shows your product with a model in different scenes, and Avatar Motion animates a person from a single photo.
Frequently asked questions
What makes a good input image?
A sharp image at the aspect ratio you want the video in, with the subject clearly visible. Small text and busy edges tend to warp once things start moving.
What is the difference between image to video and reference to video?
Image to video usually starts the clip from your picture. Reference to video uses your images as a guide to who or what should appear, so the first frame can be completely new.
How is pricing calculated?
Mostly per run, based on the resolution and length you choose. Each model page lists its prices, and some partner models report the exact cost when the run finishes.
Can I generate many videos at once?
Yes. How many run at the same time depends on your account balance, and above $250 there is no limit. See concurrency limits.