HappyHorse 1.1: video and sound generated together

Alibaba's video model turns a prompt, a still image, or up to nine reference images into 1080p clips, with dialogue, ambient sound, music, and Foley produced in the same pass, and lip-sync that holds across seven languages.

Generate a Clip

Inside HappyHorse 1.1

  • Three ways to start a scene

    Generate from a text prompt, animate a still as your first frame, or drive a scene from 1 to 9 reference images. One model handles text-to-video, image-to-video, and reference-to-video.

  • Sound made with the picture

    Dialogue, ambient noise, music, and Foley are generated in the same pass as the visuals, not layered on afterward, so a clip arrives already scored and mixed to the action.

  • Lip-sync in seven languages

    Phoneme-level lip-sync matches the spoken audio across English, Mandarin, Cantonese, Japanese, Korean, German, and French. Put the lines in your prompt and the mouth follows.

  • Characters that stay on-model

    Reference up to nine subjects as character1 through character9 and they hold their look across shots, so the same face and wardrobe carry through a sequence.

  • Multi-shot in a single prompt

    Sequence several shots in one generation by leading each segment with a timecode range, like 00-05 then 05-10, each with its own action and framing.

  • The specs that matter

    720p or 1080p, 24 fps, 3 to 15 seconds per clip, aspect ratios from 16:9 to 9:16 and beyond. Cleared for commercial use.

What is HappyHorse 1.1?

HappyHorse 1.1 is Alibaba's AI video model that generates video and synchronized audio together in a single pass. It works from a text prompt, a still image, or up to nine reference images, producing 720p or 1080p clips of 3 to 15 seconds with native dialogue. On Artlist, you can generate with HappyHorse 1.1 and use the output commercially.

AI-generated horse created with HappyHorse 1.1 on Artlist

How to prompt HappyHorse 1.1?

HappyHorse 1.1 follows scene direction closely, so the more you specify, the more control you keep. Spell out motion, framing, lighting, and pacing rather than leaving them to the model. A few patterns get the most out of it:

  1. 01

    Write the lines, hear them spoken

    For talking-head shots, write the spoken lines directly in the prompt. The audio and lip-sync are generated to match them.

  2. 02

    Sequence shots with timecodes

    To sequence multiple shots in one clip, lead each segment with a timecode range, like 00-05 then 05-10, and give each its own action.

  3. 03

    Name your characters, lock their look

    In reference-to-video, name your subjects character1 through character9 in upload order so they stay consistent across shots.

  4. 04

    Lock identity with clean references

    Use clean, high-resolution reference images with a single clear subject to lock identity and wardrobe.

Learn how to create with HappyHorse

Tutorials and launch notes to help you get more out of the HappyHorse model family.

Frequently asked questions

HappyHorse 1.1 is Alibaba's June 2026 update to HappyHorse 1.0. It keeps the same foundation: video and audio generated together in one pass, phoneme-level lip-sync in seven languages, and 3 to 15 second clips at 720p or 1080p. The update brings more dynamic motion, steadier character consistency between generations, and improved visual quality.

On Artlist, HappyHorse 1.1 also adds reference-to-video, so you can drive a scene from up to nine reference images and keep the same characters across shots. HappyHorse 1.0 works from text or a single image only.

HappyHorse 1.1 is Alibaba's AI video model that generates video and synchronized audio in a single pass. It runs text-to-video, image-to-video, and reference-to-video through one pipeline, producing 720p or 1080p clips of 3 to 15 seconds with dialogue, ambient sound, music, and Foley.

Yes. HappyHorse 1.1 synthesizes audio jointly with the picture in the same pass, so dialogue, ambient noise, music, and Foley arrive built into the clip. It also produces phoneme-level lip-sync across seven languages: English, Mandarin, Cantonese, Japanese, Korean, German, and French.

HappyHorse 1.1 outputs 720p or 1080p at 24 fps, with clips from 3 to 15 seconds and a default length of 5 seconds. There is no native 4K, and longer sequences are built by stitching multiple generations together.

Yes. HappyHorse 1.1 is cleared for commercial use. Confirm how Artlist's plan and license terms apply to your specific project before publishing, since commercial scope and any usage limits are set at the platform level.

Still have questions? We're here to help.