About Image to Video

Image to Video is a guide to turning a still image into video: which model does what, how to prompt for motion, how to prepare the source image, and what you are allowed to do with the result. Every capability claim carries the date it was checked, because they change monthly.

It is published by the Image to Video editorial desk, and its pages are written and edited by the Image to Video desk.

Independent guide. Not affiliated with Google, OpenAI, Runway, Kuaishou, ByteDance, MiniMax, Alibaba or any model provider, and we sell no credits, plans or courses. Model capabilities change monthly: every claim here carries the date it was true.

How these pages are written

Each page begins as a research brief: what the page has to answer, and which sources can answer it. The draft that follows is produced with the help of large language models, and is then put through mechanical checks that reject a figure with no named source, a claim of first-hand testing nobody did, and a price that would be wrong within the year. A page that fails those checks is rewritten rather than published with a caveat.

That process is the reason for the house rules you will notice while reading: numbers arrive with the source and the year attached, nothing here claims to have tested anything in person, and where a fact could not be sourced the page says so instead of guessing.

Corrections

If something here is wrong - a figure, a date, a link that no longer resolves - write to [email protected] with the address of the page. Corrections are made on the page itself rather than noted at the bottom of it.