Happy Horse 1.1 is Alibaba's answer to short, sound-first video generation. It always generates audio with the picture, which means voices, music, and ambient sound are synchronized by design. If you need natural-sounding short videos with sound built in, Happy Horse 1.1 is the model.

Why creators choose Happy Horse 1.1

Most video models treat audio as optional or an afterthought. Happy Horse 1.1 generates it from the start, so sound and picture sync perfectly without extra passes or dubbing steps. This changes the workflow: you describe what you want to hear and see, and Happy Horse 1.1 makes it all at once.

Native audio generation in one pass

Audio and video come out together, already synchronized. No separate dubbing step, no lipsync corrections, no out-of-sync guessing. The model handles it in one generation, which saves hours on every clip. For social content and shorts, this speed and built-in sync matter.

Short and punchy clips with character

A 3 to 10 second clip is the sweet spot for social content and short-form pieces. Happy Horse 1.1 packs motion, character, and sound into that window without bloat. The length constraint forces focus, which works well for TikTok, Reels, and YouTube Shorts.

Smooth motion with multi-character support

Happy Horse 1.1 handles multiple characters in a single clip. The model keeps faces consistent and motion smooth across characters, so a dialogue or multi-person scene stays coherent across all 10 seconds.

Lipsync and dialogue without extra steps

Character dialogue stays in sync with mouth movement automatically. Describe what you want each character to say, and the model generates matching motion and audio together.

Six aspect ratios for any format

Support for 21:9, 16:9, 4:3, 1:1, 3:4, and 9:16 means one brief can output for widescreen monitors, YouTube, TikTok, and Instagram Stories all at once. No separate generations for different platforms.

How to use Happy Horse 1.1 in Craze

Happy Horse 1.1 is built for creators who want sound baked in. Describe the video and any reference images, Craze handles the rest.

1. Start a project and describe the video. Include the tone of voice or sound you want. Add reference images if you need specific characters or looks to appear.
2. Pick Happy Horse 1.1 in the Models panel, or name it in chat. Craze shows the cost before planning.
3. Craze plans the clip, writes the prompt with your references, and generates video and sound together.
4. Review in the editor. Regenerate any shot that misses. The audio is already synced, so playback works immediately.

Specific references help. A character sheet, product shot, or location photo keeps the visuals consistent. Sound direction helps too: "upbeat narrator voice" is more useful than just "make it sound good."

Happy Horse 1.1 pricing

You pay $0.148 per second of video. The cost depends on resolution (720p or 1080p) and clip length. $0.148 per second of video covers both video and audio generation in the same price.

Use the cost calculator on this page to estimate any clip before you generate.

Who Happy Horse 1.1 is for

Social media shorts with personality

TikTok, Instagram Reels, and YouTube Shorts benefit from personality and sound. Happy Horse 1.1 delivers both in a short clip. The native audio and lipsync make social content feel polished without extra editing.

Product announcements with voiceover

A 10-second product reveal with a spokesperson voice is one generation. Reference images keep the product consistent. The voiceover syncs with motion automatically, so no re-dubbing if you want to adjust pacing.

Short-form branded spots and explainers

From hook to close in under 10 seconds, with sound that matches the mood. Happy Horse 1.1 builds short-form content fast. The built-in audio means you don't need a separate sound-design step.

Lip-synced character dialogue

The model generates dialogue and visuals in sync, which is useful for character-driven shorts or simple animated conversations. Two characters talking works as one generation, which saves time compared to generating video first and dubbing later.

Happy Horse 1.1 vs other models

Happy Horse 1.1 is built for short clips (3-10 seconds) with always-on native audio. Wan 3.0 goes up to 30 seconds, offers optional audio, and supports reference images and first/last frames. Choose Happy Horse 1.1 for punchy, sound-first content under 10 seconds. Choose Wan 3.0 for longer pieces, silent b-roll, or multi-shot sequences. For even longer content with more reference control, Seedance 2.5 runs up to 30 seconds.