MiniMax H3 Max

Model ID

Use with the Workers AI binding, the AI Gateway REST API, or the Unified API.

Calling the provider directly? Use the provider's model name instead, including with provider-native endpoints or the provider's own SDK. For example, Anthropic's name for anthropic/claude-opus-4.8 is claude-opus-4-8.

Third-party

A fast multimodal video generation model supporting text-to-video and image-to-video generation at 480P and 768P.

Model Info
Terms and License	link ↗
More information	link ↗
Pricing	Default (per second)$0.05 @480p (per second)$0.05 @768p (per second)$0.08

Usage

const response = await env.AI.run(
  'minimax/h3-max',
  {
    content: [
      {
        type: 'text',
        text: 'A golden retriever runs along a beach at sunrise while the camera tracks beside it in a cinematic slow motion shot.',
      },
    ],
    duration: 5,
    extra: { prompt_expansion_mode: 'balanced' },
    ratio: '16:9',
    resolution: '768P',
  },
)
console.log(response)

curl https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/ai/run \
  --header "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  --header "Content-Type: application/json" \
  --data '{
  "model": "minimax/h3-max",
  "input": {
    "content": [
      {
        "type": "text",
        "text": "A golden retriever runs along a beach at sunrise while the camera tracks beside it in a cinematic slow motion shot."
      }
    ],
    "duration": 5,
    "extra": {
      "prompt_expansion_mode": "balanced"
    },
    "ratio": "16:9",
    "resolution": "768P"
  }
}'

{
  "task": {
    "id": "441805798383681",
    "model": "MiniMax-H3-Max",
    "status": "succeeded",
    "created_at": 1789415547,
    "updated_at": 1789415554,
    "content": {
      "url": "https://examples.aig.cloudflare.com/minimax/h3-max/balanced-text-to-video.mp4"
    },
    "resolution": "768P",
    "duration": 5,
    "usage": {
      "total_seconds": 5,
      "input_seconds": 0,
      "output_seconds": 5,
      "input_image_count": 0
    },
    "ratio": "16:9",
    "task_type": "generation"
  }
}

Examples

Disabled Prompt Expansion — Generate a fast 480P vertical social video with prompt expansion disabled.

const response = await env.AI.run(
  'minimax/h3-max',
  {
    content: [
      {
        type: 'text',
        text: 'A chef tosses vegetables in a wok in a bright street-food kitchen, energetic handheld camera.',
      },
    ],
    duration: 5,
    extra: { prompt_expansion_mode: 'disabled' },
    ratio: '9:16',
    resolution: '480P',
  },
)
console.log(response)

curl https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/ai/run \
  --header "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  --header "Content-Type: application/json" \
  --data '{
  "model": "minimax/h3-max",
  "input": {
    "content": [
      {
        "type": "text",
        "text": "A chef tosses vegetables in a wok in a bright street-food kitchen, energetic handheld camera."
      }
    ],
    "duration": 5,
    "extra": {
      "prompt_expansion_mode": "disabled"
    },
    "ratio": "9:16",
    "resolution": "480P"
  }
}'

{
  "task": {
    "id": "441804654764158",
    "model": "MiniMax-H3-Max",
    "status": "succeeded",
    "created_at": 1789415577,
    "updated_at": 1789415583,
    "content": {
      "url": "https://examples.aig.cloudflare.com/minimax/h3-max/disabled-prompt-expansion.mp4"
    },
    "resolution": "480P",
    "duration": 5,
    "usage": {
      "total_seconds": 5,
      "input_seconds": 0,
      "output_seconds": 5,
      "input_image_count": 0
    },
    "ratio": "9:16",
    "task_type": "generation"
  }
}

Quality Prompt Expansion — Prioritize prompt expansion quality for a detailed 768P scene.

const response = await env.AI.run(
  'minimax/h3-max',
  {
    content: [
      {
        type: 'text',
        text: 'A detailed miniature railway travels through a glowing underground crystal cavern, with tiny passengers looking out of the windows.',
      },
    ],
    duration: 8,
    extra: { prompt_expansion_mode: 'quality' },
    ratio: '21:9',
    resolution: '768P',
  },
)
console.log(response)

curl https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/ai/run \
  --header "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  --header "Content-Type: application/json" \
  --data '{
  "model": "minimax/h3-max",
  "input": {
    "content": [
      {
        "type": "text",
        "text": "A detailed miniature railway travels through a glowing underground crystal cavern, with tiny passengers looking out of the windows."
      }
    ],
    "duration": 8,
    "extra": {
      "prompt_expansion_mode": "quality"
    },
    "ratio": "21:9",
    "resolution": "768P"
  }
}'

{
  "task": {
    "id": "441805071421550",
    "model": "MiniMax-H3-Max",
    "status": "succeeded",
    "created_at": 1789415588,
    "updated_at": 1789415659,
    "content": {
      "url": "https://examples.aig.cloudflare.com/minimax/h3-max/quality-prompt-expansion.mp4"
    },
    "resolution": "768P",
    "duration": 8,
    "usage": {
      "total_seconds": 8,
      "input_seconds": 0,
      "output_seconds": 8,
      "input_image_count": 0
    },
    "ratio": "21:9",
    "task_type": "generation"
  }
}

H3 Max Image to Video — Animate a supplied first-frame image at 480P.

const response = await env.AI.run(
  'minimax/h3-max',
  {
    content: [
      {
        type: 'text',
        text: 'The camera slowly pans across the scene while fabric and hair move naturally in the wind.',
      },
      {
        type: 'image_url',
        image_url: {
          url: 'https://filecdn.minimax.chat/public/85c96368-6ead-4eae-af9c-116be878eac3.png',
        },
        role: 'first_frame',
      },
    ],
    duration: 5,
    extra: { prompt_expansion_mode: 'balanced' },
    ratio: 'adaptive',
    resolution: '480P',
  },
)
console.log(response)

curl https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/ai/run \
  --header "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  --header "Content-Type: application/json" \
  --data '{
  "model": "minimax/h3-max",
  "input": {
    "content": [
      {
        "type": "text",
        "text": "The camera slowly pans across the scene while fabric and hair move naturally in the wind."
      },
      {
        "type": "image_url",
        "image_url": {
          "url": "https://filecdn.minimax.chat/public/85c96368-6ead-4eae-af9c-116be878eac3.png"
        },
        "role": "first_frame"
      }
    ],
    "duration": 5,
    "extra": {
      "prompt_expansion_mode": "balanced"
    },
    "ratio": "adaptive",
    "resolution": "480P"
  }
}'

{
  "task": {
    "id": "441804606931065",
    "model": "MiniMax-H3-Max",
    "status": "succeeded",
    "created_at": 1789415661,
    "updated_at": 1789415668,
    "content": {
      "url": "https://examples.aig.cloudflare.com/minimax/h3-max/h3-max-image-to-video.mp4"
    },
    "resolution": "480P",
    "duration": 5,
    "usage": {
      "total_seconds": 5,
      "input_seconds": 0,
      "output_seconds": 5,
      "input_image_count": 1
    },
    "ratio": "adaptive",
    "task_type": "generation"
  }
}

H3 Max Reference Images — Use multiple reference images to guide a fast character generation.

const response = await env.AI.run(
  'minimax/h3-max',
  {
    content: [
      {
        type: 'text',
        text: 'Create a short cinematic portrait of the character walking through a modern gallery, preserving the face and clothing from the references.',
      },
      {
        type: 'image_url',
        image_url: {
          url: 'https://cdn.hailuoai.com/prod/hailuo_demo/testsets/h3_promo_eval_ref2va/gallery/sr_v2p26_trio_seed42_20260724/inputs/9d5c7a33fa6e_01_%E5%9B%BE1_MHGgbVga3o_gpt4o-image-1780651118146.png',
        },
        role: 'reference_image',
      },
      {
        type: 'image_url',
        image_url: {
          url: 'https://cdn.hailuoai.com/prod/hailuo_demo/testsets/h3_promo_eval_ref2va/gallery/sr_v2p26_trio_seed42_20260724/inputs/7af326902315_00_%E5%9B%BE2_YqtKbY1jpo_u4391985813_Young_male_wearing_cream_hoodie_and_dark_brown_sh_45d56ed0-d626-4c37-9a5b-77f51f374982_1.png',
        },
        role: 'reference_image',
      },
    ],
    duration: 5,
    extra: { prompt_expansion_mode: 'balanced' },
    resolution: '768P',
    ratio: '16:9',
  },
)
console.log(response)

curl https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/ai/run \
  --header "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  --header "Content-Type: application/json" \
  --data '{
  "model": "minimax/h3-max",
  "input": {
    "content": [
      {
        "type": "text",
        "text": "Create a short cinematic portrait of the character walking through a modern gallery, preserving the face and clothing from the references."
      },
      {
        "type": "image_url",
        "image_url": {
          "url": "https://cdn.hailuoai.com/prod/hailuo_demo/testsets/h3_promo_eval_ref2va/gallery/sr_v2p26_trio_seed42_20260724/inputs/9d5c7a33fa6e_01_%E5%9B%BE1_MHGgbVga3o_gpt4o-image-1780651118146.png"
        },
        "role": "reference_image"
      },
      {
        "type": "image_url",
        "image_url": {
          "url": "https://cdn.hailuoai.com/prod/hailuo_demo/testsets/h3_promo_eval_ref2va/gallery/sr_v2p26_trio_seed42_20260724/inputs/7af326902315_00_%E5%9B%BE2_YqtKbY1jpo_u4391985813_Young_male_wearing_cream_hoodie_and_dark_brown_sh_45d56ed0-d626-4c37-9a5b-77f51f374982_1.png"
        },
        "role": "reference_image"
      }
    ],
    "duration": 5,
    "extra": {
      "prompt_expansion_mode": "balanced"
    },
    "resolution": "768P",
    "ratio": "16:9"
  }
}'

{
  "task": {
    "id": "441805100810489",
    "model": "MiniMax-H3-Max",
    "status": "succeeded",
    "created_at": 1789415671,
    "updated_at": 1789415685,
    "content": {
      "url": "https://examples.aig.cloudflare.com/minimax/h3-max/h3-max-reference-images.mp4"
    },
    "resolution": "768P",
    "duration": 5,
    "usage": {
      "total_seconds": 5,
      "input_seconds": 0,
      "output_seconds": 5,
      "input_image_count": 2
    },
    "ratio": "16:9",
    "task_type": "generation"
  }
}

Parameters

▶content[]

arrayrequiredminItems: 1

duration

integerrequiredminimum: 5maximum: 15

ratio

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

callback_url

stringformat: uri

resolution

stringrequiredenum: 480P, 768P

▶extra{}

object

▶task{}

object

API Schemas (Raw)

Input

Output