Create cutscenesfor your game
Describe the shot or start from a frame of your game, and render it with frontier video models.
Alibaba
HappyHorse 1.1 for game videos
HappyHorse 1.1 is an Alibaba video model with synchronized native audio and multilingual lip sync, so spoken lines can go straight into the prompt. It placed in the top ten for both text to video and image to video on the Artificial Analysis video arena in September 2026. In Summer it renders 720p clips from a prompt, a still or up to nine reference images.
Made with HappyHorse 1.1
She draws the katana in one fluid motion and spins into a slow-motion sweeping cut; rain droplets burst away from the blade and her hair and sash swirl; she finishes in a poised stance as the camera circles slowly around her. One continuous shot, no cuts, consistent character, correct anatomy and hands, no text, no logos, no captions.
A cheerful cartoon fox girl adventurer in a green scarf skips across a sunny meadow toward the camera, waves and laughs, soft 3D animated film style, gentle camera push in
She turns her head toward the camera, the falcon on her shoulder ruffles its feathers, a light breeze moves her braid, slow push in
What game makers use it for
- Talking NPC and narrator clips with lines written in the prompt.
- Character intros for trailers in several languages.
- Animating a portrait that speaks one line of dialogue.
In Summer
- Price
- 21 credits per second of video
- Inputs
- Text to video, Image to video, Reference to video