Video models
Models that make a new video from words, pictures and clips.
Wan 3Studio's everyday video model — video from words, from a first and last frame, from your pictures, clips and sound, or a change to a clip you have.
Wan 3 PrimeThe same model as Wan 3 on a faster footing — it answers sooner and costs more; it is not a finer picture.
Seedance 2.5Video from words, or from a large set of your own pictures, a clip and a sound — with a whole multi-shot film written into one take.
Grok Imagine Video 1.5Short clips from your words and your reference pictures, with one reference sound beside them if you want.
HappyHorse 1.1Clips from words, from one first frame, or from one to several of your reference pictures — with a seed to bring a take back.
Kling 3.0 ProThe higher edition of the pair — video from your words, or a first frame and a last one brought to life, with sound if you want it.
Kling 3.0 StandardThe everyday edition of the pair — video from words in three shapes, or a first and last frame brought to life.
LTX-2.5 FastThe fast edition — video from words or from a first and last frame, in wide or tall, with sound if you want it.
MiniMax H3Clips at one large size, from your words in shapes from wide cinema to tall, or from a first and last frame that keep their own shape.
MiniMax Hailuo 2.3Silent clips from your words or from one first frame — no shape, size or seed to choose; your words go as written.
PixVerse V6Clips of whole seconds from your words, or from a first and last frame — with sound, and a seed to bring a take back.
Veo 3.1Video from your words, or from a first and last frame, in wide or tall — and it can make the clip's own sound.
Veo 3.1 FastThe quicker, cheaper edition of the same model — video from words or frames, wide or tall, with its own sound if you want it.