WebExtrator 智能抽取 API 集成指南 - Ace Data Cloud

POST https://api.acedata.cloud/webextrator/extract 在 /webextrator/render 的基礎上做內容抽取。除了 render 介面的全部參數，額外支援：

欄位	類型	必填	預設	說明
`expected_type`	string	❌	`markdown`	期望抽取產物：`markdown` / `article` / `text` / `links` / `structured`
`enable_llm`	boolean	❌	false	啟用 LLM 後處理（適合 `article` / `structured`）
`instruction`	string	❌	-	LLM 抽取指令，例如「抽取商品標題、價格、規格」

同步回應

{
  "success": true,
  "task_id": "550e8400-...",
  "trace_id": "550e8400-...",
  "started_at": "2026-05-02T10:30:00.123Z",
  "finished_at": "2026-05-02T10:30:08.789Z",
  "elapsed": 8.666,
  "data": {
    "kind": "extract",
    "expected_type": "article",
    "url": "https://example.com/post/1",
    "title": "示例文章",
    "author": "張三",
    "published_at": "2026-05-01",
    "content": "# 示例文章\n\n正文 ...",
    "summary": "本文介紹 ..."
  }
}

非同步模式、錯誤碼、計費規則與 /webextrator/render 完全一致。

範例：抽取文章正文（啟用 LLM）

curl -X POST https://api.acedata.cloud/webextrator/extract \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://example.com/news/1",
    "expected_type": "article",
    "enable_llm": true
  }'

範例：非同步 + 自訂結構化抽取

curl -X POST https://api.acedata.cloud/webextrator/extract \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://shop.example.com/item/123",
    "expected_type": "structured",
    "enable_llm": true,
    "instruction": "抽取商品標題、價格、庫存、3 張主圖 URL",
    "callback_url": "https://your-domain.com/wbx-callback"
  }'