Any article, clean structured JSON.
Point it at a news story, blog post, or any long-form page and get back the title, author, publish date, clean body text, images, and feeds. Deterministic, multi-language, no LLM bill.
{
"article": {
"title": "Why we should learn German",
"author": "John le Carré",
"pub_date": "2017-07-02 00:05:12",
"text": "I began learning German at the age of 13...",
"html": "<p>I began learning German...</p>",
"image": "https://.../lead.jpg",
"images": ["https://.../fig-1.jpg"],
"language": "en",
"site_name": "the Guardian",
"is_article": 1
},
"credits_used": 5
}See it work, before you sign up.
Drop in a URL, run a real call against the live API, and watch the JSON come back in about a second. No API key required.
Nav, ads, and boilerplate stripped out. Get plain text and sanitised HTML of just the article.
Byline, publish and modified dates, canonical URL, site name, favicon, language - all detected for you.
Lead image plus in-article images (filtered by size + relevance), embedded videos, and RSS feeds.
Non-Latin scripts and mixed encodings handled cleanly. Accurate extraction in 40+ languages.
Flip on headless Chrome for JS-heavy pages, rotate residential proxies, and solve CAPTCHAs on the hard targets.
Stitch multi-page articles into one result, or pass raw_html you already have and skip the fetch entirely.
Copy. Paste. Ship.
curl 'api.ujeebu.com/extract?url=https://example.com/post&media=1&images=1' \
-H 'ApiKey: YOUR_API_KEY'import { Ujeebu } from 'ujeebu';
const uj = new Ujeebu(process.env.UJEEBU_KEY);
const res = await uj.extract({
url: 'https://example.com/post',
media: true,
images: true,
});
console.log(res.article.title, res.article.author);from ujeebu import Ujeebu
uj = Ujeebu(api_key='YOUR_API_KEY')
res = uj.extract(
url='https://example.com/post',
media=True,
images=True,
)
print(res['article']['title'], res['article']['author'])Real things real teams shipped this quarter.
Turn thousands of article URLs into a clean, searchable text corpus with author and date attached. Track coverage without babysitting parsers.
Pull full-text from blogs, publishers, and newsrooms into your own reader, newsletter, or knowledge base, images and feeds included.
Deterministic, boilerplate-free article text is the cleanest input for embeddings and retrieval. Pair with the Markdown API for chunk-ready output.
5 credits per Article Extractor call. Here's what that buys.
One credit pool covers every endpoint. Failed calls cost 0. No per-feature upcharges, no premium-proxy tax. See full pricing →
Next up: any page to JSON, automatically.
Article extraction is what we do today, deterministically, for blogs, news, and long-form pages. Auto-Extract takes the same idea to every other page type: point it at a product listing, a profile, a directory row, and it detects the page type and returns structured JSON, with no selectors and no schema to write.
- Any page type. Products, listings, profiles, and more, not just articles.
- No rules, no schema. It works out the shape of the page and hands back clean JSON.
- Same endpoint you already call. A single type=auto flag on /extract, when it ships.
On the roadmap. The /extract endpoint is built to grow into it, so today's article integration keeps working unchanged.
{
"page_type": "product",
"data": {
"name": "Sony WH-1000XM5",
"price": 399.99,
"currency": "USD",
"in_stock": true,
"rating": 4.7
},
"credits_used": 20
}Ship Article Extractor tonight.
5,000 credits free. No card. Real residential proxies on the free tier.