Dumpling AI 每月有 18,752 次造訪,月增率 +27.8%,定價模式為免費增值。PulpMiner 每月有 1,756 次造訪,成長率 -9.2%,定價模式為付費。
PulpMiner turns any public webpage into a live JSON API. Paste a URL, let AI read the page structure and propose a clean JSON shape, adjust the fields you want, then save it as an endpoint that returns fresh data on every call. It handles JavaScript-rendered pages, supports dynamic URL variables so one endpoint serves many pages, and plugs into Google Sheets, Zapier, n8n and MCP-compatible AI assistants — no scraper code to write or maintain.
| 面向/指標 | ![]() | |
|---|---|---|
| 官方網域 | dumplingai.com | pulpminer.com |
| 定價模式 | 免費增值 | 付費 |
| 每月造訪次數 | 1.88萬 (18,752) | 1756 (1,756) |
| 月流量變化量 | +4078 (+27.8%) | -178 (-9.2%) |
| 主要市場與市佔 | 🇺🇸 美國 (25.8%) | 🌐 全球 (—) |
| 每月直接造訪 | 7636 | 684 |
| 使用者評分 | 4.1 | 4.1 |
| 支援平台 | 網站 | 網站 |

為 AI 代理打造的抓取、搜尋與資料擴充 API
每項能力自動路由到上游優質供應商,需要精確輸出時可固定 firecrawl.scrape、serper.search 等原生端點。
從靜態網站、完整渲染應用與整站爬取中取得乾淨頁面內容,支援 markdown 等格式,無須維護脆弱爬蟲。
透過統一驗證層進行網頁、新聞、地圖與自動完成查詢,並取得 YouTube/TikTok 字幕、影片中介資料、LinkedIn 資料與 Google 評論。
以 doc-to-text 與結構化擷取介面,將 PDF 與文件轉為乾淨文字或結構化欄位,供下游代理流程使用。
PulpMiner turns any public webpage into a live JSON API. Paste a URL, let AI read the page structure and propose a clean JSON shape, adjust the fields you want, then save it as an endpoint that returns fresh data on every call. It handles JavaScript-rendered pages, supports dynamic URL variables so one endpoint serves many pages, and plugs into Google Sheets, Zapier, n8n and MCP-compatible AI assistants — no scraper code to write or maintain.
Paste any public URL and PulpMiner proposes a structured JSON shape you can adjust, then saves it as a persistent API endpoint returning fresh data on every call.
The extractor reads page structure and pulls exactly the fields you asked for — price, title, author, stock status and so on — as clean JSON without writing selectors.
Make parts of the URL dynamic so one endpoint serves many pages through parameters, and turn on optional caching for pages you hit often to save credits.
Renders JavaScript, so single-page apps and dynamically loaded content are supported alongside plain HTML pages.
免費 / 開源用提示詞取代選擇器,把任何網站變成結構化 JSON
免費 / 開源內建代理、瀏覽器與反爬處理的網頁擷取 API
免費 / 開源一次 API 呼叫,從任意網站擷取乾淨的結構化資料