AI Models

SnapVivid runs every model behind one studio and one credit balance, so you never juggle separate tools or logins. Leave the model on Auto and SnapVivid routes your prompt to the best available engine, or pick a specific model when you want a particular look. New models are added as providers come online — your prompts and library carry over unchanged.

The models split into two families. Video models — Seedance, and soon Veo, Kling, Sora and Wan — turn a photo or a written scene into motion, each with its own strengths in camera movement, realism and how faithfully it holds your subject. Image models, led by Google's Nano Banana, generate and edit stills, keeping a face or a product consistent while you restyle, relight or recompose it. Open any model's page for what it does best, or just start generating and switch models mid-project without losing your prompt or your library.

Auto-routing is the default because the best model is rarely the same for every shot: a portrait, a landscape and a product turn each reward a different engine. Leave the selector on Auto to let SnapVivid choose, or pick a model by name the moment you want a specific look — the credit cost and the controls stay the same either way.