Develop video ideas from text, a starting image, or multiple references, with audio-supported generation across the three Model Studio modes.
Use Model Studio’s happyhorse-1.1-t2v for a written scene. Name the participants and central action, such as a shopkeeper sliding a gift toward a customer. The mode has its own ID; the R2V reference-image allowance does not apply.
Model Studio’s happyhorse-1.1-i2v uses a first-frame image and a text prompt. The image establishes the opening composition; describe the subject’s next action and the camera movement to develop the shot.
Model Studio’s happyhorse-1.1-r2v accepts 1–9 reference images. Identify their roles in the prompt, such as appearance, clothing or setting. Each image can be JPEG, PNG or WebP, up to 20 MB, with a shortest side of at least 400 pixels.
Model Studio lists audio support for HappyHorse 1.1 text-to-video, first-frame image-to-video and reference-to-video. Review the generated sound with the picture, checking important dialogue, audible events and the timing of visible actions.
| Dimension | HappyHorse 1.1 | Veo 3.1 | Wan 3.0 Prime |
|---|---|---|---|
| Input formats | Text, image, multi-image reference modes | Text and image; visual references | Text, image, video, audio; task-specific combinations |
| Core direction | Separate text, image, and reference-to-video modes | Video generation with audio and visual guidance | Multi-shot video with multimodal references |
| Audio capabilities | Audio-supported T2V, I2V and R2V | Native audio generation | Native dialogue, music, and effects |
| Video duration limits | Model Studio: 3–15 seconds in T2V, I2V and R2V | Gemini API: 4, 6, or 8s initially; references and 1080p/4K require 8s | Model Studio: 2–30s; video inputs have a combined duration limit |
| Output resolution | Model Studio: 480P, 720P, 1080P in all three modes | Gemini API: 720p; 1080p/4K at 8s; extension at 720p | Model Studio: 480P, 720P, 1080P |
| Suggested uses | Character interactions planned from text, images, or references | Scenes planned around sound, visual references, and continuation | Multi-shot stories and reference-led scene development |
Specifications checked October 19, 2026. Sizes and controls refer to the named provider and task. Sources: Model Studio video model specifications · HappyHorse first-frame API · HappyHorse reference-image API · Google Veo 3.1 API guide · Wan Prime model guide · Wan 3.0 generation API · Wan 3.0 generation guide
Use text or an input image to establish a person and opening scene, then describe the action and setting changes. This suits clips built around a character moment; review faces, limbs and motion closely.
Assign a clear role to each reference image, such as appearance, clothing or setting. This suits video proposals with established visual references; avoid conflicting instructions across those references.
Establish the opening with a chosen image, then describe the intended action and sound. This suits content that develops a still into an audiovisual moment, with sound cues that support the visible action.
Explore related video workflows for character scenes, visual references, and still-image motion.
The Model Studio happyhorse-1.1-r2v API accepts 1–9 images in JPEG, JPG, PNG or WebP format. Each must be at most 20 MB with a shortest side of at least 400 pixels. Use [Image 1], [Image 2] and so on to identify each reference in the prompt.
The happyhorse-1.1-i2v API is a first-frame workflow: provide one opening image and a text description of the movement that follows. Model Studio documents first/last-frame guidance separately for Wan 3.0; it is a different input workflow.
Model Studio lists integer durations from 3 to 15 seconds for T2V, I2V and R2V. All three modes support 480P, 720P and 1080P output at 24 fps in MP4 format.
Model Studio lists audio support for text-to-video, first-frame image-to-video and reference-to-video. Listen to the complete output alongside the picture and check speech, audible events and timing before using the clip.
The improvements described by the provider are not a guarantee for every output. A character can remain recognizable in the opening yet change during a turn or interaction. Review the full clip against the identity details required by your particular scene.
