MiniMax H3 API - Omni-Modal AI Video Generation in 2K

MiniMax/minimax-h3

MiniMax H3 is a general-purpose omni-modal generation model with unified understanding of text, image, video, and audio context, native stereo audio-video output, strong instruction following, and high-quality 2K video generation.

Output: $0.12 / use or 8 uses / $1

Input

Output

{
  "task_id": "j2n94b12anrmw0cwdf1bpjde6r",
  "user_id": 1,
  "version": "c98270c13de8ee8e86597b77ec3493956f1bd8b8ea4d0021bb0641541babbd3b",
  "error": null,
  "total_time": 240,
  "predict_time": 240,
  "logs": null,
  "output": [
    "https://vmodel.ai/data/dev/model/minimax/minimax-h3/121061af-e563-4bc7-9b58-b9bddc2ff9b3.mp4"
  ],
  "status": "succeeded",
  "create_at": 1771334410,
  "completed_at": 1771334960,
  "input": {
    "text": "A cinematic product shot of a silver sports car driving through a neon-lit city at night, reflections moving across the bodywork, dynamic tracking camera.",
    "resolution": "768P",
    "duration": 5,
    "ratio": "16:9"
  }
}

Generated in: 240 seconds

Download

Input

Output

{
  "task_id": "j2n94b12anrmw0cwdf1bpjde6r",
  "user_id": 1,
  "version": "c98270c13de8ee8e86597b77ec3493956f1bd8b8ea4d0021bb0641541babbd3b",
  "error": null,
  "total_time": 240,
  "predict_time": 240,
  "logs": null,
  "output": [
    "https://vmodel.ai/data/dev/model/minimax/minimax-h3/121061af-e563-4bc7-9b58-b9bddc2ff9b3.mp4"
  ],
  "status": "succeeded",
  "create_at": 1771334410,
  "completed_at": 1771334960,
  "input": {
    "text": "A cinematic product shot of a silver sports car driving through a neon-lit city at night, reflections moving across the bodywork, dynamic tracking camera.",
    "resolution": "768P",
    "duration": 5,
    "ratio": "16:9"
  }
}

Generated in: 240 seconds

Download

HTTP Request

Run minimax/minimax-h3:c98270c13de8ee8e86597b77ec3493956f1bd8b8ea4d0021bb0641541babbd3b using Vmodel's HTTP API.

  curl -X POST https://api.vmodel.ai/api/tasks/v1/create
    -H "Authorization: Bearer $VModel_API_TOKEN"
    -H "Content-Type: application/json"
    -d '{
    "version": "c98270c13de8ee8e86597b77ec3493956f1bd8b8ea4d0021bb0641541babbd3b",
    "input": {}
}'

Input Schema

The fields you can use to run this model with an API. If you don't give a value for a field its default value will be used.

text

Type: string

Default value: -

Description: Required text description of the video to generate.

first_frame

Type: image

Default value:

Description: Optional first-frame image used to guide the opening frame and visual content.

last_frame

Type: image

Default value:

Description: Optional last-frame image used to guide the ending frame and visual content.

resolution

Type: enum

Default value: -

Description: Generated video resolution.

Choices: 768P, 2K

duration

Type: int

Default value: -

Description: Generated video duration in seconds. Enter an integer from 4 through 5.

Range: Min: 4 | Max: 15

ratio

Type: enum

Default value: adaptive

Description: Generated video aspect ratio. Adaptive selects the most suitable ratio from the input; the actual ratio is available in the query response's ratio field.

Choices: adaptive, 21:9, 16:9, 4:3, 1:1, 3:4, 9:16

Examples

Pricing

Model pricing for minimax/minimax-h3. Looking for volume pricing? Get in touch.

When

⚙ model variant is 768p

$0.120000

per generation at 768P

or 8 generations for $1

When

⚙ model variant is 2k

$0.180000

per use

or 5 uses for $1

Readme