Magick API Official Developer Documentation

Welcome to Magick API. This page, based on the organized source documentation, provides the shortest integration path to help you complete your first call.

Tip: You only need to replace the domain and API KEY with Magick API's to migrate seamlessly with zero code changes!

• Domain: "https://api.magickapi.com", fully compatible with OpenAI-style interfaces
• API KEY: Generate in the dashboard

Step 1: Obtain an API Key

1. Open the API Key Management page.
2. Log in to your account and create a new API Key.
3. Keep the key secure. Pass it in subsequent calls via Authorization: Bearer YOUR_API_KEY.

Step 2: Choose an Interface Type

Text and most image interfaces typically return results synchronously; video interfaces use an asynchronous task model, returning id / task_id upon successful submission.

Step 3: Send a Request

Text Generation Example

curl -X POST https://api.magickapi.com/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5",
    "messages": [
      {
        "role": "user",
        "content": "帮我调查并推荐一个最适合编码的大模型"
      }
    ]
  }'

Image Generation Example

curl -X POST https://api.magickapi.com/v1/images/generations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-image-2",
    "prompt": "可爱的猫咪坐在劳斯莱斯后座上哭",
    "size": "1:1",
    "n": 1
  }'

Video Generation Example

curl -X POST https://api.magickapi.com/v1/videos \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "veo-3.1-fast-generate-preview",
    "prompt": "一个小女孩坐在窗台欣赏春风吹动的柳树",
    "duration": 8,
    "size": "1280x720"
  }'

Step 4: Query Task Status

After the video interface returns id / task_id, poll it through the OpenAI-compatible video task endpoint:

curl -X GET https://api.magickapi.com/v1/videos/YOUR_TASK_ID \
  -H "Authorization: Bearer YOUR_API_KEY"

After the task is completed, download the video content:

curl -L https://api.magickapi.com/v1/videos/YOUR_TASK_ID/content \
  -H "Authorization: Bearer YOUR_API_KEY" \
  --output output.mp4

It is recommended to wait 5 to 10 seconds after submission before starting to poll, then query every 3 to 5 seconds.

Next Steps

• Text conversation interface: Start with General Chat Interface (Default Streaming).
• OpenAI Responses: Read OpenAI Responses Multimodal Interface.
• Image generation: Read GPT-Image-2 Image Generation.
• Video generation: Read VEO3 Video Generation.