Hydrafetch

Give it domains and get each site's design as tokens: colors by role, fonts, type scale, radius, shadows and spacing, plus a paste-ready CSS stylesheet and Tailwind CSS v4 theme to restyle an app like it. Input: domains like linear.app. One row per site; failures are not charged.

Give it company names, optionally with the country after a comma, and get each company's official website and domain, checked against the company's own homepage so directories never slip through, with a confidence level. One row per company; names we cannot confirm are not charged.

Give it domains and get a full company record for each: name, legal name, tagline, description, logo, brand colors, fonts, industry with NAICS code, email and social profiles. Input: domains like stripe.com. One row per company; domains that return nothing are not charged.

Give it page URLs and find out whether LLMs and AI agents can read them: a score out of 100, every check passed, warned or failed, the specific fixes, and a sample of the markdown an agent would see. Input: full URLs. One row per URL; pages we cannot read are not charged.

Search Amazon.com with terms like "wireless earbuds", or give it product links or ASINs, and get each product's US price, rating, review count, sales in the past month and seller, with the full product page on request. One row per product; searches with no products are not charged.

Give it App Store or Google Play links, Apple app ids or package names and get app details plus its newest reviews: rating, rating count, installs, version, then each review's text, stars, date and developer reply. One row per review; apps not found or with no reviews are not charged.

Give it article or blog post URLs and get the full text with its title, author, publish date, site name, language, lead image and word count. Input: full URLs. One row per article; pages with no article to extract are not charged.

Give it domains and get a direct image URL for each company's logo, with its format, width, height, dominant color and whether it is an icon or a wordmark. Input: domains like stripe.com. One row per company; domains with no logo found are not charged.

Give it company domains and get the ads each company runs on Google, LinkedIn and Facebook and Instagram in one dataset with the same columns: platform, headline, copy, call to action, image or video, link and dates. One row per ad; domains with no ads anywhere are not charged.

Give it company domains and get every open job from the hiring platform each company uses: title, team, location, remote status, employment type, pay when listed, posting date, apply link and full description. Input: domains like stripe.com. Companies with no open roles found are not charged.

Give it brand names, domains, Facebook pages or Ad Library links and get the ads each runs on Facebook and Instagram: copy, images, videos, carousel cards, platforms, start and end dates and active status. One row per ad; brands with no ads, or none we can match, are not charged.

Give it company domains and get the Google ads each company runs, from the public Ads Transparency Center: advertiser, format, image, preview link and first and last shown dates. Input: domains like stripe.com, plus how many ads per domain. One row per ad; domains with no ads are not charged.

Search Google Maps with a phrase like "dentists in Berlin" or a Maps link and get every place: name, address, phone, website, rating, reviews and coordinates, up to 120 per search. Turn on contacts to add emails and socials from each place's own website, charged only when found.

Give it search terms and get Google Trends data: interest over time with a 0 to 100 value per point, interest by country or region, and top and rising related queries. Input: one search per line, up to 5 comma-separated terms to compare. Searches with no data are not charged.

Give it company names, domains or LinkedIn pages and get the ads each runs on LinkedIn: full copy, headline, call to action, image or video, run dates, impressions by country and targeting. One row per ad; companies with no ads, or none we can match, are not charged.

Give it LinkedIn company page URLs or company domains and get the public profile: industry, company size, employees on LinkedIn, followers, headquarters address, founded year, specialties and logo. No login needed. One row per company; companies we cannot find are not charged.

Search LinkedIn jobs with terms like "data engineer" or a jobs search link, filtered by location, date posted, workplace, job type and experience, and get each posting with company, pay, applicants, seniority and full description. No login needed. Searches with no jobs are not charged.

Search the news for any topic, like "stripe funding", and get recent articles with their full text as clean markdown, the real publisher link, source, author and publish date. Input: search queries, plus how many articles and how far back. Queries with no readable articles are not charged.

Give it links to PDF files and get each document as clean markdown with headings and tables intact, plus its title and word count, ready for LLMs, RAG and search. Input: full URLs to PDFs. One row per PDF; files we cannot read are not charged.

Give it Shopify store domains and get every product: title, price range, currency, variants with SKUs and stock, options, images, vendor, type, tags and description, including stores that hide their product feed. One row per product; stores with no readable products are not charged.

Give it company domains and get each company's Trustpilot TrustScore, star breakdown and total reviews, plus its newest reviews with rating, text, dates and the company's reply, without reviewer names. Input: domains like stripe.com. One row per review; companies not on Trustpilot are not charged.

Give it a start URL and get the site or docs section crawled into clean markdown, one row per page with its title, language, page type, depth and word count. Input: full URLs, plus how many pages to read per site. Charged per page returned; pages that return nothing are free.

Give it URLs and get a screenshot of each page as a hosted PNG link, of the first screen or the full page, rendered in a real browser, with its width and height. Input: full URLs. One row per screenshot; pages that fail to load are not charged.

Give it URLs and get each page as clean markdown for LLMs, RAG and AI agents, with navigation, cookie banners and ads removed and JavaScript pages rendered. Input: full URLs. Each row has the markdown, title, language and word count; pages that return nothing are not charged.