About Reference to Video

Reference to Video is built around a simple idea: creators should be able to direct a sequence without losing the people, products, outfits, movement, or visual language that made the first shot work.

Why reference-first video matters

Generic image-to-video tools often treat every generation as a fresh start. That can be useful for a single clip, but it creates friction when a story needs the same character in a Tokyo street, a café, and a rooftop—or when a product must remain recognizable across an entire campaign. Faces drift, clothes change, props move between characters, and style fragments from shot to shot.

Reference to Video makes references part of the creative brief. You can combine image, video, and audio inputs, describe the next shot, and choose a model suited to the job. The goal is not to promise mathematically perfect continuity. It is to give creators a clearer, more controllable workflow for producing connected shots.

Who it is for

• Short-drama and story creators building recurring characters
• UGC and ecommerce teams producing repeatable product ads
• Anime, VTuber, and original-character creators
• Fashion and AI-influencer teams protecting outfit identity
• Music-video creators matching a performer, movement, and sound
• Brand and game teams working with mascots or stylized characters

How we work

We show real use cases, explain which reference types each model accepts, and keep unsupported controls out of the interface. Seedance 2.0 is available for demanding multi-reference work, while faster options are available when iteration speed matters. Model capabilities can change, so the generator only exposes settings that the selected model can actually use.

Questions, corrections, or partnership enquiries are welcome at contact@referencetovideo.net.