Happy Horse 1.0 video generation and reference-guided editing

Happy Horse 1.0 is an Alibaba video family on Vidgo with text, image, reference and video-edit tasks. New generation supports 3–15 seconds and 720p/1080p. Editing accepts a 3–60-second source, optional target images and audio handling controls, with its own detected-duration billing.

Prepare a written scene, a starting frame, separate subject references or a source clip. Describe the action, environment and visual details to retain, then generate or edit with the matching task. Reviewed results can support nature scenes, character performances, architectural animation, restoration proposals, material-style studies and weather or lighting variations.

Sign in to Vidgo and review your balance and API credentials in the account dashboard. Test the selected task online, then follow its documentation to submit the model identifier, input fields and output settings. Retain the returned task ID to query status and retrieve the completed video, or configure a callback for your application.

Choose text, one starting image or one to nine references for a new 3–15-second video. Choose Video Edit for a 3–60-second source clip and optional one-to-five target images. Select 720p or 1080p; text/reference generation also selects a supported ratio, while editing selects auto or origin audio handling.

New generation offers whole-number durations from 3 to 15 seconds and 720p/1080p. Text and reference tasks offer 16:9, 9:16, 1:1, 4:3 and 3:4; image generation uses its source frame. Editing uses the source clip’s duration, selected resolution and audio_setting auto or origin. Prompts accept up to 2,500 characters; an image-animation prompt is optional. seed accepts an integer from 0 to 2,147,483,647.

Use public, directly downloadable HTTP(S) URLs. Image generation uses one image_urls entry; reference generation uses one to nine reference_image_urls. Editing requires one video_url for a 3–60-second clip, no larger than 100 MiB for duration inspection, and optionally one to five reference images. Explain each image’s role in the prompt and prepare media consistent with the target scene.

New video generation is billed by the requested 3–15-second duration and the selected resolution rate. Video editing uses its separate rate multiplied by the detected source-video duration, capped at 15 billable seconds. A 3–60-second source remains the editing input; the billing cap describes the charge calculation. The dynamic table shows each task’s current rate and unit. Open the task with the intended settings and source clip to review the complete estimate before submitting.

Choose 1.0 when the project uses its text, image, reference or existing-video editing workflow, including source-audio preservation and target-image guidance for edits. Happy Horse 1.1 provides its own three generation tasks, nine text/reference framing options, image Base64 support and rates. Match the actual media and controls to your project and compare complete estimates and test results.

Identify the action or visible content to change and name the face, materials, architecture, light and camera motion to retain. Assign each reference a clear role in defining the subject, setting or target appearance. For video editing, describe the target scene and use optional images to guide its appearance; select origin when the existing audio should remain. Inspect the actual output against the clip and references.