MiniMax H3

Vendor: MiniMax

Video

Create high-detail videos from text, or from image, video, and audio references

Typical latency: 60-240s
Reference inputs: Up to 15
toolId: minimax-h3

Input modes

text-to-videoimage-to-videoomni-reference

Variants

Standard

Create a new 768P or 2K video with text, frames, or multimodal references

text-to-videoimage-to-videoomni-reference

Fast

Generate faster with H3 Max at 480P or 768P

text-to-videoimage-to-video

768P → 2K

Create an additional 2K version from an eligible H3 768P result

video-upscale

Parameters

Parameter	Type	Default	Values	Notes
duration	select	5	4 \| 5 \| 6 \| 8 \| 10 \| 12 \| 15	—
resolution	select	768p	768p \| 2k	—
aspect_ratio	select	9:16	21:9 \| 16:9 \| 4:3 \| 1:1 \| 3:4 \| 9:16	—

Pricing

Pricing will be published shortly — contact us for early access rates

Prices last updated: —

API example

POST /v1/generations — minimax-h3

curl -X POST https://api.mukelabs.com/v1/generations \
  -H "Authorization: Bearer $MUKE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "toolId": "minimax-h3",
    "inputMode": "text-to-video",
    "prompt": "your prompt here",
    "params": { "duration": "4" },
    "references": [],
    "idempotencyKey": "a-unique-key-per-task",
    "subjectId": "your-user-id"
  }'

FAQ

How do I call MiniMax H3 on Muke Labs?

Submit POST /v1/generations with toolId "minimax-h3" and one of its input modes (text-to-video, image-to-video, omni-reference), then poll GET /v1/generations/{jobId} for the result.

What does MiniMax H3 cost?

Public pricing for this model is being finalized — contact us for early access rates. All tasks are billed from your prepaid USD balance and failed tasks refund automatically.

What happens if a task fails?

The job returns status "failed" with a typed error code, and its charge is returned to your balance automatically.

Where do I get an API key?

Create one in the developer portal (https://portal.muke.name), then send it as a Bearer token on every request.