HappyHorse

Video GenerationReference GuidedVideo Editing

Alibaba’s HappyHorse 1.0 combines video generation, reference-guided creation, and instruction-based editing in one APIXO route. Build 3–15-second videos from text or images, guide new scenes with multiple visual references, or transform an existing video while choosing 720p or 1080p output and workflow-specific audio behavior.

APIXO Workflows

Generate / Reference / Edit

APIXO Price

From $0.125/sec720p unit price

APIXO Resolution

720p / 1080p

Generation Length

3–15 seconds

Audio

Included by default

Loading workspace...

HappyHorse Model API Rate Structure

Text-to-Video

Resolution

720p

APIXO price

$0.125 / second

Text-to-Video

Resolution

1080p

APIXO price

$0.225 / second

Image-to-Video

Resolution

720p

APIXO price

$0.125 / second

Image-to-Video

Resolution

1080p

APIXO price

$0.225 / second

Reference-to-Video

Resolution

720p

APIXO price

$0.125 / second

Reference-to-Video

Resolution

1080p

APIXO price

$0.225 / second

Video Edit

Resolution

720p

APIXO price

$0.125 / second

Video Edit

Resolution

1080p

APIXO price

$0.225 / second

Mode	Resolution	APIXO price
Text-to-Video	720p	$0.125 / second
Text-to-Video	1080p	$0.225 / second
Image-to-Video	720p	$0.125 / second
Image-to-Video	1080p	$0.225 / second
Reference-to-Video	720p	$0.125 / second
Reference-to-Video	1080p	$0.225 / second
Video Edit	720p	$0.125 / second
Video Edit	1080p	$0.225 / second

Checking API access...

Integrate with AI

Copy a prompt for your AI assistant to install APIXO MCP, read the live schema, and wire this model into your app.

Endpoints

Create task

POSThttps://api.apixo.ai/api/v1/generateTask/happyHorse

Poll status

GEThttps://api.apixo.ai/api/v1/statusTask/happyHorse?taskId={taskId}

Request body

request_type

async

Use callback with callback_url for webhooks.

input

object

Mode, prompt, references, size, duration, and audio settings.

moderequired

Typeselect

Choose whether HappyHorse should create a clip from text, animate one image, follow multiple references, or edit an existing video.text-to-video · image-to-video · reference-to-video · video-edit

promptoptional

Typestring

Required for text-to-video, reference-to-video, and video-edit; optional for image-to-video. Describe the scene, subject, motion, camera language, style, and audio cues.

resolutionrequired

Typeselect

720p is the lowest-cost production option. Choose 1080p when higher output detail matters.720p · 1080p

durationrequired

Typeselect

Output length for text-to-video, image-to-video, and reference-to-video. Video-edit billing uses input video duration instead.3 · 5 · 10 · 15Only in mode=text-to-video | image-to-video | reference-to-video

ratioconditional

Typeselect

Supported for text-to-video and reference-to-video. Image-to-video follows the uploaded first frame.16:9 · 9:16 · 1:1 · 4:3 · 3:4Only in mode=text-to-video | reference-to-video

audio_settingconditional

Typeselect

Video-edit only. Auto lets HappyHorse infer audio, while origin keeps source-video audio when supported.auto · originOnly in mode=video-edit

watermarkoptional

Typeboolean

Add or remove a watermark on the generated output.

seedoptional

Typetext

Optional random seed from 0 to 2147483647 for repeatable generations.

Parameter	Type	Required	Description
`mode`	select	required	Choose whether HappyHorse should create a clip from text, animate one image, follow multiple references, or edit an existing video.text-to-video · image-to-video · reference-to-video · video-edit
`prompt`	string	optional	Required for text-to-video, reference-to-video, and video-edit; optional for image-to-video. Describe the scene, subject, motion, camera language, style, and audio cues.
`resolution`	select	required	720p is the lowest-cost production option. Choose 1080p when higher output detail matters.720p · 1080p
`duration`	select	required	Output length for text-to-video, image-to-video, and reference-to-video. Video-edit billing uses input video duration instead.3 · 5 · 10 · 15Only in mode=text-to-video \| image-to-video \| reference-to-video
`ratio`	select	conditional	Supported for text-to-video and reference-to-video. Image-to-video follows the uploaded first frame.16:9 · 9:16 · 1:1 · 4:3 · 3:4Only in mode=text-to-video \| reference-to-video
`audio_setting`	select	conditional	Video-edit only. Auto lets HappyHorse infer audio, while origin keeps source-video audio when supported.auto · originOnly in mode=video-edit
`watermark`	boolean	optional	Add or remove a watermark on the generated output.
`seed`	text	optional	Optional random seed from 0 to 2147483647 for repeatable generations.

File parameters (URL arrays)

image_urls[mode=image-to-video | reference-to-video | video-edit]Image-to-video requires exactly one first-frame image. Reference-to-video supports 1–9 reference images. Video edit accepts up to 5 optional reference images.Only in mode=image-to-video | reference-to-video | video-edit

video_urls[mode=video-edit]Upload exactly one source video for video-edit mode.Only in mode=video-edit

Example (cURL)

curl -X POST "https://api.apixo.ai/api/v1/generateTask/happyHorse" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "request_type": "async",
    "input": {
      "mode": "text-to-video",
      "prompt": "a cinematic rainy night in tokyo",
      "resolution": "1080p",
      "ratio": "16:9",
      "duration": 5,
      "watermark": false
    }
  }'

Full API documentation

Capabilities

Four distinct ways to direct a video

Text-to-Video Creation

Generate a new clip from a required prompt describing subjects, action, camera language, atmosphere, style, and audio cues. APIXO exposes five aspect ratios for this workflow, including horizontal, vertical, square, and 4:3 framing.

Fixed-Frame Animation

Image-to-video accepts exactly one source image and uses it as the first frame. The prompt is optional, allowing either open-ended animation or more explicit direction for movement, camera behavior, and sound.

Multi-Reference Guidance

Reference-to-video combines a required prompt with one to nine images. These assets guide subjects and visual appearance within a newly generated scene; they are references rather than fixed opening or closing frames.

Instruction-Based Editing

Video-edit transforms one existing video according to a required instruction. Add up to five optional image references when the edit needs visual guidance for a subject, object, environment, or stylistic direction.

Integrated Audio Output

HappyHorse output includes audio by default through APIXO. Generation modes do not expose an audio-disable switch, while video-edit can use automatic audio behavior or preserve the source video’s original audio.

Seed and Watermark Controls

APIXO exposes 720p and 1080p output, an optional random seed, and a watermark setting. Resolution affects the billing rate, while seed control supports more structured experimentation across related requests.

Workflow Groups

Choose new generation or source-based transformation

Generate a New Video

Use text-to-video for prompt-only creation, image-to-video when one image must become the first frame, or reference-to-video when multiple images should guide a new composition. APIXO applies the 3–15-second duration control to these three workflows.

Text and Images

Edit an Existing Video

Supply exactly one input video, a required edit instruction, and optionally up to five reference images. Video-edit does not use APIXO’s public generation-duration field; output length and billing are derived from the source video instead.

Video Transformation

API Specifications

HappyHorse controls through APIXO

Production limits vary by workflow and should be applied only where documented.

Up to 2,500 characters

Prompt Length

16:9 / 9:16 / 1:1 / 4:3 / 3:4

Generation Ratios

1 fixed first frame

Image-to-Video

1–9 images

Reference Guidance

1 video + 0–5 images

Edit Inputs

MP4 video URL

APIXO Result

Production Uses

Build adaptable short-form video pipelines

Ecommerce

Product Advertising Variations

Generate product demonstrations and campaign concepts from text, animate an approved product still as the opening frame, or use several reference views to guide a new scene. Review product geometry, labels, and claims before publishing.

Digital Presenters

Presenter and Dialogue Concepts

Develop short presenter, spokesperson, and dialogue-led concepts with integrated audio output. Reference images can guide appearance, but exact identity, speech content, pronunciation, and lip synchronization should be validated for every production language and subject.

Brand Production

Reference-Guided Campaign Scenes

Combine character, product, wardrobe, environment, or style references in a single generation request. Unlike first-frame animation, the supplied images act as visual guidance and do not prescribe the opening composition.

Postproduction

Guided Video Restyling

Transform one existing clip with instructions for appearance, environment, or visual treatment. Preserve the original soundtrack when continuity matters, or select automatic audio behavior when the edited result should follow HappyHorse’s default handling.

Workflow Notes

Apply the correct rules to each mode

APIXO constraints

Text-to-video, reference-to-video, and video-edit require prompts; image-to-video permits an omitted prompt.

Generation modes accept 3–15-second duration; video-edit does not expose a public duration parameter.

Aspect-ratio selection applies only to text-to-video and reference-to-video through APIXO.

Generated audio cannot be fully disabled through the published APIXO schema.

APIXO supports asynchronous polling or webhook delivery and returns one generated video URL.

Inputs and outputs remain subject to provider safety review; important results should be stored promptly.

HappyHorse questions

The documented workflows correspond to Alibaba Cloud Model Studio’s HappyHorse 1.0 family: happyhorse-1.0-t2v, happyhorse-1.0-i2v, happyhorse-1.0-r2v, and happyhorse-1.0-video-edit. APIXO groups these workflows under its happyHorse route.

Only in image-to-video mode, which requires exactly one image and uses it as the first frame. Reference-to-video accepts one to nine images as visual guidance for a new scene rather than fixing them as video frames.

APIXO charges $0.125 per second at 720p and $0.225 per second at 1080p. Text-to-video, image-to-video, and reference-to-video are billed from the selected duration at the chosen resolution rate.

APIXO bills video-edit as input-video seconds plus generated-output seconds, multiplied by the selected resolution rate. The billable input side is capped at 15 seconds, and the generated side is billed at the same capped duration.

Yes. In video-edit mode, set APIXO’s audio_setting to origin to preserve the input video’s original audio. The alternative auto setting uses the route’s automatic audio behavior. No fully silent setting is publicly exposed.