Wan API, AI Video Generation API

Headers Parameters:

{
  "api-key": "Your API Key",
}

Submit Generation Task

Path: /wan/tasks

Method: POST

Type: application/json

Body:

{
  "gen_params": {
    "prompt": "A young woman with a red umbrella walking in the rain",
    "steps": 16,
    "duration": 3,
    "size": "480*832",
    "image": read_local_img_as_b64
  },
  "model": "14b-i2v-480p"
}

Description:

• model: Model type, available options:
  • 1.3b-t2v-480p: Small model text-to-video (480P) Coming Soon
  • 14b-t2v-480p: Large model text-to-video (480P) Coming Soon
  • 14b-i2v-480p: Large model image-to-video (480P) Available
  • 14b-t2v-720p: Large model text-to-video (720P) Coming Soon
  • 14b-i2v-720p: Large model image-to-video (720P) Coming Soon
• gen_params: Generation parameters
  • prompt: Prompt description, supports multiple languages
  • steps: Generation steps, range 14-36, higher steps consume more USD. Recommended value: 18.
  • duration: Video duration, available options: 3 seconds, 5 seconds, 8 seconds (not yet available)
  • size: Video size, supports multiple ratios (1:1, 9:16, 16:9, 3:4, 4:3), available options include:
    • 480P: 512*512, 640*640, 480*832, 832*480, 540*720, 720*540
    • 720P: 720*1280, 1280*720, 960*960, 832*1088, 1088*832 (not yet available)
  • image: Required only for image-to-video models, base64 encoded image data

Response:

{
    "status": "completed",
    "task_id": "0e7a0481-ebca-4ebb-9dcf-b9bc74147338",
    "gen_params": {
        "prompt": "A young woman with a red umbrella walking in the rain",
        "steps": 16,
        "duration": 3,
        "size": "480*832",
        "r3_key": "wan-images/0e7a0481-ebca-4ebb-9dcf-b9bc74147338.jpg"
    },
    "data": {
        "end_time": "2025-03-13T01:59:52",
        "video_url": "https://cdn2.foxai.me/wan-video/task_id.mp4"
    },
    "model": "wan-14b-i2v-480p",
    "created_at": "2025-03-13T01:59:52"
}

Status Changes

Upon task creation, the status is pending, then it is picked up by the GPU and changes to queued, then starts processing and changes to processing, upon completion, it changes to completed, and if it fails, it changes to failed.

Pricing Rules

The Wan API pricing is calculated based on the model type, generation steps, and video duration.

Base Price (By Model Type)

Steps Coefficient

Duration Coefficient

Pricing Formula

Final Cost (USD) = Base Price × Steps Coefficient × Duration Coefficient

Pricing Examples

1. Using 14b-i2v-480p model, 18 steps, 3-second video:

   • 0.2(base) × 1.0(steps) × 0.8(duration) = $0.16
   • Takes approximately 2 minutes to generate.

2. Using 14b-t2v-720p model, 25 steps, 5-second video:

   • 0.25(base) × 1.25(steps) × 1.0(duration) = $0.3125

Query

Query Task Details

Path: /wan/task

Method: GET

Parameters:

• ids: Task IDs to query, multiple IDs separated by commas, up to 25 tasks per query

Example Request:

GET /api/v1/wan/task?ids=0e7a0481-ebca-4ebb-9dcf-b9bc74147338,1f8a0582-fbda-5fcc-0ecf-c0cd85258449

Response:

[
  {
    "status": "completed",
    "task_id": "0e7a0481-ebca-4ebb-9dcf-b9bc74147338",
    "gen_params": {
      "prompt": "A young woman with a red umbrella walking in the rain",
      "steps": 16,
      "duration": 3,
      "size": "480*832",
      "r3_key": "wan-images/0e7a0481-ebca-4ebb-9dcf-b9bc74147338.jpg"
    },
    "data": {
      "end_time": "2025-03-13T01:59:52",
      "video_url": "https://cdn2.foxai.me/wan-video/0e7a0481-ebca-4ebb-9dcf-b9bc74147338.mp4"
    },
    "model": "14b-i2v-480p",
    "created_at": "2025-03-13T01:59:52"
  },
  {
    "status": "processing",
    "task_id": "1f8a0582-fbda-5fcc-0ecf-c0cd85258449",
    "gen_params": {
      "prompt": "A colorful hot air balloon flying over mountains",
      "steps": 20,
      "duration": 5,
      "size": "512*512"
    },
    "data": {},
    "model": "14b-t2v-480p",
    "created_at": "2025-03-13T02:05:12"
  }
]

Description:

• The interface returns an array containing details of each requested task
• If a task ID does not exist or is inaccessible, it will not appear in the results
• Possible status values:
  • pending: Waiting to be processed
  • queued: In the queue
  • processing: Being processed
  • completed: Processing completed
  • failed: Processing failed

Video Hosting

FoxAI users own all rights to the videos they generate, and the user who generated the video is solely responsible for it.

We recommend users to download and host the videos themselves.