UniAll AI SEO/GEO · Happy Horse Edit · 2026-05-31

What is Happy Horse Edit API?

Happy Horse Edit API lets developers and creative teams generate or edit short videos with prompts and media inputs. On UniAll AI, the public model id is `happy-horse-edit-720p`. It supports asynchronous video generation through `/v1/videos/generations`, with modes for source video editing, image-to-video, and reference-media-driven video creation.

The model is useful when you want to automate video variations, product clips, ad creatives, social media assets, or visual concept tests without building a custom video pipeline from scratch.

Core capabilities

Happy Horse Edit supports three practical workflows:

• **Video edit**: provide a source video URL and a prompt describing the edit.
• **Image to video**: provide 1–5 reference images and a prompt to create motion from visual references.
• **Reference media to video**: create a short video guided by prompt and reference assets.

Supported output duration is **3–15 seconds**. Output resolution can be **720p** or **1080p**. The model returns one video per generation request.

Who should use it?

Happy Horse Edit is a good fit for:

• **Marketing teams** creating short ad variations and product reveal clips.
• **E-commerce teams** turning product photos or existing videos into motion assets.
• **Developers** building AI video tools, batch creative systems, or internal content workflows.
• **Agencies and platforms** that need API-based video editing with predictable parameters.
• **Social media teams** producing short-form video素材 for campaigns, tests, and storyboards.

API usage overview

Endpoint:

```http POST /v1/videos/generations ```

The API is asynchronous, so your application should submit a generation task and then poll or handle task status according to your UniAll AI integration pattern.

Example request body:

```json { "model": "happy-horse-edit-720p", "generation_mode": "video_to_video", "prompt": "Make the product look premium with cinematic lighting, smooth motion, and a clean studio background.", "video_url": "https://example.com/source.mp4", "duration": 5, "resolution": "720p", "video_count": 1 } ```

For image-to-video generation, use `generation_mode: "image_to_video"` and pass `reference_image_urls`:

```json { "model": "happy-horse-edit-720p", "generation_mode": "image_to_video", "prompt": "Animate the product with a slow camera orbit and soft studio lighting.", "reference_image_urls": ["https://example.com/product.png"], "duration": 5, "resolution": "1080p", "video_count": 1 } ```

Key parameters

| Parameter | Type | Notes | |---|---:|---| | `model` | string | Use `happy-horse-edit-720p` for the public 720p model id. | | `generation_mode` | string | Use `video_to_video`, `image_to_video`, or `reference_media_to_video`. | | `prompt` | string | Required. Describe the visual edit, motion, scene, or style. | | `video_url` | string | Required for video editing mode. Supports common video inputs such as MP4, MOV, and WebM. | | `reference_image_urls` | array | Required for image-to-video mode. Supports 1–5 PNG, JPEG, or WebP images. | | `duration` | integer | 3–15 seconds. Default is 5 seconds. | | `resolution` | string | `720p` or `1080p`. | | `video_count` | integer | Fixed at 1 video per generation. |

Pricing angle

Happy Horse Edit is billed per generated second. On UniAll AI user pricing, 720p starts at **$0.1904 per second**, while 1080p is **$0.3808 per second**. Because cost scales with duration and resolution, teams should use 720p for drafts and 1080p for final assets when higher detail is required.

A simple cost estimate is:

```text estimated cost = duration in seconds × resolution unit price ```

For example, a 5-second 720p generation is estimated from 5 × $0.1904 before any account-specific pricing rules or platform adjustments.

Best use cases

• Create product display videos from existing product footage.
• Generate short ad concepts from reference images.
• Re-style source videos with a prompt-based creative direction.
• Produce test clips for landing pages, social ads, or campaign variants.
• Build automated video workflows for internal tools or customer-facing applications.

Prompting tips

For better results, describe the intended edit clearly. Include subject, motion, lighting, style, and constraints.

Good prompt pattern:

```text Keep the original product shape and color. Add a smooth cinematic camera push-in, premium studio lighting, subtle reflections, and a clean dark background. Avoid changing the logo. ```

For video editing, mention what should stay unchanged and what should be transformed. For product and brand content, explicitly preserve logos, labels, and key object identity when needed.