MiniMax/minimax-h3-max-image-to-video/Released Aug 2026

MiniMax H3 Max: Image to Video

Prompt-optimized image-to-video generation

Commercial use

Video

About MiniMax H3 Max: Image to Video

MiniMax H3 Max (Image to Video) is a post-trained variant of MiniMax H3, tuned by fal for stronger prompt adherence and better aesthetics. It animates an optional first-frame image from a text prompt into a video, and routes to text-to-video when no image is given.

It generates 5 to 15 second clips at 480P or 768P resolution, with prompt expansion modes that trade latency for prompt fidelity, an optional end frame for first-to-last keyframe generation, and a safety checker enabled by default. When a first frame is provided, the output keeps its aspect ratio.

See what this model can do

A handful of real prompts and their outputs.

Animated Painting

Watercolor Heron

"Keeps the watercolor texture of the source still while the bird lifts off in fluid flight."

Animated Painting

Portrait Comes Alive

"Subtle, believable motion (blinks, a glance and a smile) with the painting's chiaroscuro preserved."

Food

Syrup Pour

"Glossy liquid motion and melting butter with a slow food-commercial push-in."

Archival Revival

1920s Street Scene

"Animates many subjects at once while keeping the sepia tone and old-film grain."

Pets

Windowsill Cat

"Natural animal motion from a single still, with sunlit dust and billowing curtains."

Adventure

Ridge in the Wind

"Strong environmental motion, blowing snow and drifting clouds, around a steady subject."

Animated Illustration

Poster Seaplane

"Animates a flat lithograph-style poster while keeping its colors and print texture."

Wildlife

Snowy Owl Takeoff

"Realistic wildlife motion from a single still, from a head turn to a full takeoff."

Food

Ramen Lift

"Steam, dripping broth and a slow push-in turn a food photo into a commercial shot."

Ready to build with MiniMax H3 Max: Image to Video?

Try MiniMax H3 Max: Image to Video in the Workbench to prompt it, compare outputs, and iterate on prompts without writing any code. When you're ready to ship, call the same model from our API and build your own apps on top of it.

curl -sSf -X POST https://hub.oxen.ai/api/ai/videos/generate \
    -H "Content-Type: application/json" \
    -H "Authorization: Bearer $OXEN_API_KEY" \
    -d '{
  "model": "minimax-h3-max-image-to-video",
  "prompt": "A cat sitting on a chair",
  "input_image": "https://hub.oxen.ai/api/repos/ox/Oxen-AI-Assets/file/main/images/ox_zoom_out_1926_1076.png",
  "duration": 5,
  "resolution": "768P",
  "enable_safety_checker": true,
  "sync_mode": false,
  "prompt_expansion_mode": "balanced"
}'

Request parameters

Every field MiniMax H3 Max: Image to Video accepts in the request body. See the API reference for response formats.

POSThttps://hub.oxen.ai/api/ai/videos/generate

Field	Type	Required	Default	Description
`prompt`	`string`	Yes	—	Text prompt for video generation
`input_image`	`string` uri · nullable	Yes	—	Optional URL of the image to use as the first frame. When provided, the output aspect ratio follows this image. When omitted, the request is handled as text-to-video (16:9 by default).
`tail_image_url`	`string` uri	No	—	Optional URL of the image to use as the last frame, for first-to-last keyframe generation.
`duration`	`integer`	No	`5`	The duration of the video in seconds. Range: 5 – 15
`resolution`	`string`	No	`"768P"`	The native generation resolution of the video. `480P768P`
`enable_safety_checker`	`boolean`	No	`true`	If set to true, the safety checker will be enabled.
`sync_mode`	`boolean`	No	`false`	Return the generated video as base64 instead of a CDN URL.
`prompt_expansion_mode`	`string`	No	`"balanced"`	How much effort to spend rewriting the prompt before generation. 'balanced' returns in about a second. 'quality' spends up to ~30s on a richer prompt. `balancedquality`
`seed`	`integer` nullable	No	—	Random seed. A random seed is selected when omitted.

Model JSON Video generation reference Async queue Models endpoint