Create videos from text or images with HappyHorse 1.0, with generated audio.
Direct a short scene with HappyHorse 1.0: describe the lighting, camera angle and action. Start from a photo to set the opening composition, and generate sound alongside the picture.
HappyHorse 1.0 supports 3–15s clips at 720p/1080p in 5 ratios; generated audio Supported; end frame Not available.
See how each example uses framing, movement or sound, then try the prompt with your own material.
A wide view establishes the room, a closer shot follows the reader, and a final close-up brings attention to the book. Watch the cuts, changing camera distance and soft window light in this published HappyHorse 1.0 example.
Platform-published demonstration · fal.ai · 1080p
1080p example published by fal.ai. View the original prompt at the source.
View original sourceDescribe the subject, light and camera movement to explore a new shot without an input photograph. Compare the result with the same prompt on another model.
A plain cobalt blue ceramic cup rests on a pale wooden table beside a window. Soft morning light, gentle steam rising, one slow camera push-in. No people, no lettering, no logos. Keep the cup shape and tabletop stable.
AIVG test recording · 5.041667s · 2026-09-06 · Current quote: 170 credits
720p · 5s · 16:9
Input: Text prompt; generated audio included.
What to check: The camera moves close enough to crop the cup at the end. Review framing across the whole shot.
Upload a photograph to establish the starting subject and composition. Describe the movement you want, then compare the first frame and the rest of the clip with your original.
A plain cobalt blue ceramic cup rests on a pale wooden table beside a window. Soft morning light, gentle steam rising, one slow camera push-in. No people, no lettering, no logos. Keep the cup shape and tabletop stable.
AIVG test recording · 5.041667s · 2026-09-06 · Current quote: 170 credits
720p · 5s · 16:9
Input: A photograph you own or have permission to use; generated audio included.
What to check: Compare the handle, steam, framing and background with the starting image. A starting frame does not guarantee preservation throughout the video.
How to
Use text for a new scene, or upload a clear photograph to set the starting image.
Specify the subject, one action and a camera movement. These examples use a blue cup and a slow push-in.
Check the exact credits, generate, then play the complete result in your creation. Reuse the inputs for another variation.
Capabilities
Example 720p / 5s: 170 credits, equivalent to $3.383 USD using the 100-credit / $1.99 USD pack. Credits are purchased in packs; check the exact quote for your settings before generating.
AIVG does not provide an API, unlimited credits or watermark removal; official account matters go to Alibaba.
Related models
| Model | Baseline credits | Highest resolution | Longest duration | Native audio | Modes |
|---|---|---|---|---|---|
| HappyHorse 1.0 | 170 credits · 720p / 5s | 1080p | 15s | Supported | Text to Video / Image to Video |
| Seedance 2.0 | 180 credits · 480p · 21:9 · audio off / 5s | 720p | 15s | Supported | Text to Video / Image to Video |
| Vidu Q3 Pro | 185 credits · 720p · audio off / 5s | 1080p | 16s | Supported | Text to Video / Image to Video |
| Runway Gen-4 Turbo | 150 credits · model output / 5s | decided by the model | 10s | Not available | Image to Video |
Commercial use follows each model's own terms — confirm them before publishing.
Watermark and export permissions depend on your plan and selected model.
When a generation job reaches a failed state, the credits charged for it are returned to your balance.
Exact quote before every run.
Create a new video from text or animate a starting photo. Choose the duration, resolution and available sound options in settings.
The 5-second 720p example uses 170 credits. Other settings have different prices; check the amount before generating.
This configuration creates a new video from text or a photograph. To change a specific existing clip, open AI video editing and choose a supported model such as Gemini Omni.
Each example includes its source and settings. Use these to compare the input and result; your own generation may differ.
Related models
Exact quote before every run.
