HappyHorse 1.1 R2V

HappyHorse 1.1 R2V generates video from a written brief and up to nine reference images on Alibaba Cloud Model Studio.

Reference images for a planned shot

HappyHorse 1.1 R2V is a reference-to-video service. It accepts text and images and returns video. Up to nine reference images can describe a subject, setting or visual style. This makes it a candidate for a product scene or character sequence where the references matter as much as the written prompt.

Preparing the input

Use clear references with a consistent subject, then describe motion, camera movement and the intended composition separately. Images that disagree about a product’s shape or a character’s appearance make the brief harder to interpret. Start with one shot and inspect the entire clip for changes in identity, geometry, labels and background objects before building a longer sequence.

Region and video billing

Model Studio publishes region-specific per-second rates. The Singapore route lists $0.07 at 480p, $0.14 at 720p and $0.18 at 1080p. Other regions have their own rates and availability. This service does not have an ordinary language-model context window or a text-output token price; those fields remain empty in the model table.

Comparing video tools

Hold the reference set, prompt, resolution and requested duration constant. Compare the number of usable clips, time to a finished result and editing work required. A pleasing first frame does not establish that the generated motion remains coherent throughout the shot.

Sources

Model documentation