MiniMax | H3 | Max | Image to Video

Video·minimax-h3·by Minimax

H3 Max Image-to-Video turns prompts and optional reference images into short AI video clips with 480P or 768P output for creative motion work.

Runtime (p50)
5m
Estimated price
From $0.025
Call the API
prediction.sh
sh
curl -X POST \
  -H "Authorization: Bearer $EACHLABS_API_KEY" \
  -H "Content-Type: application/json" \
  --data '{
    "model": "minimax-h3-max-image-to-video",
    "version": "0.0.1",
    "input": {
        "prompt": "Animate the existing meadow scene while strictly preserving the original chair, open sketchbook, flowers, landscape, lighting and composition. A gentle breeze moves through the tall grass and wildflowers, creating soft natural waves across the field. The pages of the open sketchbook lift and flutter slightly in the wind without turning completely. Distant grasses move subtly while warm sunlight shifts naturally through passing clouds. Use a very slow cinematic push-in toward the chair, realistic plant physics, subtle film grain and natural atmospheric movement.",
        "duration": 10,
        "image_url": "https://cdn-us.eachlabs.ai/defaults/8e10cf7554c34824b5214f0fc0d18700.png",
        "resolution": "768P",
        "enable_safety_checker": true,
        "prompt_expansion_mode": "balanced"
    },
    "webhook_url": ""
}' \
  https://api.eachlabs.ai/v1/prediction/

Related models

4 models
* FAQ

About MiniMax | H3 | Max | Image to Video

01 / 03

What is H3 Max Image-to-Video?

H3 Max Image-to-Video is a MiniMax model served through fal for AI video generation. It turns a text prompt, and optionally a starting or ending image, into a short video clip with 480P or 768P output. The image input can guide the first frame while the prompt controls motion and scene direction.