Crawl4AI tools

Get one Wikipedia article's short description, summary paragraph, link, thumbnail, top-level sections and last edit time.

Search Apple's App Store for iPhone and iPad apps, in Apple's order, with developer, price, rating, version and description.

Get the daily weather forecast for a place from Open-Meteo: conditions, high and low temperature, precipitation, wind, sunrise and sunset.

Get one iPhone or iPad app from Apple's App Store: developer, price, rating, categories, current version and description.

Get the latest Yahoo Finance quote for up to 10 tickers: price, change against the previous close, day and 52-week range and volume.

Search Wikipedia articles and get title, link, a snippet around the match, word count and last edit time.

People search by company domain and job title, one row per person: name, job title, LinkedIn profile, work e-mail when one is published (with its kind and status), department, seniority and decision role, the page they were found on, and the company's facts. Search by company (`where.company.domain`), by job title (`where.person.title`: words in any order, VP = Vice President, Head = Director), by department or seniority (department: marketing, sales, engineering, product, finance, hr, operations, legal, it, design, executive; seniority: owner_founder, c_level, vp, director, manager; decision_role for decision makers), by a person's name (`where.person.name`), by company country, city or industry, or by a free-text brief (`q`). `limit` is how many people (default 10, up to 100), one price a call at any limit. An empty answer is free. About 0.2 s.

List a public Facebook page's latest posts with text, time, reactions, comments and shares.

Read the items of any RSS 2.0 or Atom feed: title, link, publication time, author and summary.

Enrich a company from its domain, name or LinkedIn company page: name, description, industry, employee count (the people on record), LinkedIn company URL, logo, website technology stack, the work e-mails and phones published on its pages, its social links and mail provider. `domain` is the company's website (stripe.com); `linkedin` is its LinkedIn page link or handle (when several domains share the page, the most asked one answers and meta.also_on_this_linkedin_page lists the rest); `name` is the company's name (legal endings such as Inc or GmbH are ignored; answers only when one company clearly has the name, otherwise 404 with up to 5 candidate domains, free). A company not on record answers 404, free. About 0.2 s.

List the products a Shopify store publishes at /products.json with title, vendor, type, price range, stock, variants and tags.

Get one Product Hunt product: tagline, description, website, rating, followers, categories and its launches with scores.

Get one X (Twitter) account: name, bio, follower, following and post counts, join date, location, website and verified badge.

List today's launches on Product Hunt's home page in rank order, with tagline, vote score, topic and link.

Crawls one site level by level from a start URL, inside the site, through include and exclude filters, cut at max_pages; runs only as a job. Poll it with crawl4ai.web.scrape.job.status and read the pages (url, level, title, status, links, markdown or html) with crawl4ai.web.scrape.job.results.

Answers a question or a whole agent prompt (up to 4,000 characters) from the web's current top results, with the pages it drew on. Ask for JSON in the question, or pass a JSON Schema, and the answer is JSON of that shape. answer.source says whether the text is Google's own answer (google_ai_mode, google_ai_overview) or written from the pages (crawl4ai). answered is false, and free, when the pages hold no answer. Experimental: fields may be added.

List the latest posts of one X (Twitter) account with text, likes, reposts, replies, quotes, time and the repost flag.

Get one Zillow home: status, price and Zestimate, address, beds, baths, living area, year built, price history and photos.

Search arXiv papers and get title, authors, the start of the abstract, PDF link, submission time and primary category.

Read one X (Twitter) account's timeline through a from: search, newest first, with text, time, counts, views, language and media.

Get one public Facebook post: author, text, time, reactions by type, comments, shares, photos and videos.

Get one eBay listing: title, price, sale format, condition, stock and sold count, seller feedback, item specifics and images.

Scrape up to 25 URLs in one call; results stream back as NDJSON, one line per URL as it finishes, then one summary line with the batch's total cost.