HappyHorse 1.0 Video Edit modifies an input video guided by a reference image and a prompt. It uses the async task API — submit a task, then poll until ready.Documentation Index
Fetch the complete documentation index at: https://docs.anyfast.ai/llms.txt
Use this file to discover all available pages before exploring further.
⚠️ Native format required. Because video-edit needs both a video and a reference image at the same time, the unified flat-field shortcut cannot express it. You must use the native DashScope-style request body with input.media[].
Key capabilities
- Video editing — Modify objects, clothing, scenes inside an existing video
- Resolution — 720P / 1080P
- Watermark control — Optionally disable the bottom-right watermark
⚠️ Output duration is determined by the input video length. The duration parameter may not take effect for video-edit.
Workflow
Quick example
Parameters
| Parameter | Type | Required | Description |
|---|---|---|---|
model | string | Yes | Must be happyhorse-1.0-video-edit |
input.prompt | string | Yes | Text description of the edit |
input.media | array | Yes | Exactly 2 items: one video + one reference_image |
input.media[].type | string | Yes | video or reference_image |
input.media[].url | string | Yes | Public URL of the video or reference image |
parameters.resolution | string | No | 720P / 1080P. Default: 1080P |
parameters.watermark | boolean | No | Default true. Set false to remove watermark |
parameters.seed | integer | No | Random seed [0, 2147483647] |
Billing
Charged by resolution tier × billed duration. Ifduration is not provided, defaults to 5 seconds. Failed tasks are automatically refunded.
API Reference
Interactive playground for HappyHorse 1.0 Video Edit.
Task Query
Poll task status and retrieve the final video URL.