Website Tech Stack Detector: CMS, Hosting, Email and 360+ Technologies

Find out what any website is built with, so you can segment leads, size a market or research competitors. Paste a list of websites and get back the technology stack of each one: CMS, ecommerce platform, JavaScript frameworks, analytics, ad pixels, tag managers, CDN, hosting, web server, live chat, payment, email hosting, email delivery services and DNS provider. Every detection comes with its evidence, and versions are reported when the site exposes them.

It works from a single homepage request plus a few DNS lookups per site, so it is fast, cheap and does not need proxies or a browser.

Quick start

1. Click Try for free, paste websites into Websites and press Start.
2. Or call the API with an empty input {}. That analyzes these 3 demo sites:

{
    "urls": ["shopify.com", "wordpress.org", "vercel.com"]
}

On Nov 1, 2026 this run analyzed all 3 sites for 3 x $0.002 = $0.006. It found 18 technologies on shopify.com (Shopify, Tailwind CSS, Cloudflare, Zendesk, Google Workspace, SendGrid ...), 11 on wordpress.org (WordPress, Nginx, PHP, MySQL, Jetpack, Google Tag Manager ...) and 19 on vercel.com. The vercel.com row, with 15 of its 19 technologies cut:

{
    "input": "vercel.com",
    "domain": "vercel.com",
    "status": "ok",
    "httpStatus": 200,
    "title": "Agentic Infrastructure - Vercel",
    "technologies": [
        { "name": "Next.js", "categories": ["JavaScript framework", "Web framework"], "confidence": "detected",
          "evidence": ["header x-powered-by: Next.js, Payload", "header x-nextjs-prerender"] },
        { "name": "Vercel", "categories": ["PaaS"], "confidence": "detected",
          "evidence": ["header server: Vercel", "header x-vercel-id"] },
        { "name": "Google Workspace", "categories": ["Email hosting"], "confidence": "detected",
          "evidence": ["dns mx: aspmx.l.google.com"] },
        { "name": "Postmark", "categories": ["Email delivery"], "confidence": "detected",
          "evidence": ["dns txt: v=spf1 ... include:..."] }
    ]
}

What makes it different

• DNS-based detection. Most tech lookup tools only read the HTML. This Actor also reads MX, TXT (SPF and verification records), NS, CNAME and A records, so it finds things that never appear on the page: Google Workspace or Microsoft 365 email, SendGrid, Mailgun, Amazon SES, HubSpot, Salesforce, Zendesk, the DNS provider and the real hosting platform.
• Evidence for every result. Each technology says why it was detected (header, cookie, script URL, meta tag, DNS record), so you can trust or filter the results.
• Versions where the site reveals them (WordPress, jQuery, Bootstrap, PHP, Nginx and more).
• Failed sites are free. You only pay for websites that answered and were analyzed. Dead domains, invalid inputs, blocked sites and sites that disallow us in robots.txt cost nothing.
• Polite by default. Respects robots.txt, identifies itself with a clear User-Agent and fetches one page per site.

Use cases

• Lead generation and sales prospecting: find every Shopify store, WordPress site or HubSpot customer in a list, or split leads by stack before outreach.
• Competitor research: see which analytics, ad networks, chat tools and hosting your competitors use.
• Market research: count technology adoption across thousands of domains.
• CRM enrichment: add CMS, ecommerce platform and email provider columns to your accounts.
• Agencies and freelancers: find sites on outdated platforms or old versions that need a rebuild.
• Security and IT audits: inventory web servers, CDNs and exposed versions across your own domains.

What it detects

More than 360 technologies in 50 categories, including:

Category	Examples
CMS and website builders	WordPress, Drupal, Joomla, Ghost, Webflow, Wix, Squarespace, Framer, HubSpot CMS, Elementor
Ecommerce	Shopify, WooCommerce, Magento, BigCommerce, PrestaShop, Shopware, Salesforce Commerce Cloud
JavaScript frameworks and libraries	React, Next.js, Vue.js, Nuxt.js, Angular, Svelte, Astro, Gatsby, jQuery, Alpine.js
UI frameworks	Bootstrap, Tailwind CSS, Font Awesome
Analytics and tag managers	Google Analytics, Google Tag Manager, Hotjar, Plausible, Matomo, Mixpanel, Segment, Microsoft Clarity
Advertising	Meta Pixel, Google Ads, LinkedIn Insight Tag, TikTok Pixel, Microsoft Advertising
Marketing and CRM	HubSpot, Mailchimp, Klaviyo, Marketo, Salesforce, ActiveCampaign
Live chat and support	Intercom, Drift, Zendesk, Crisp, Tawk.to, LiveChat, Freshdesk
CDN, hosting and PaaS	Cloudflare, Fastly, Akamai, Amazon CloudFront, Vercel, Netlify, GitHub Pages, AWS, Google Cloud, Heroku
Web servers and languages	Nginx, Apache, LiteSpeed, Microsoft IIS, PHP, ASP.NET, Node.js, Express
Email hosting (DNS)	Google Workspace, Microsoft 365, Zoho Mail, Proton Mail, Fastmail
Email delivery (DNS)	SendGrid, Mailgun, Amazon SES, Postmark, Mailchimp, Brevo
DNS providers	Cloudflare DNS, Amazon Route 53, GoDaddy, Namecheap, Google Cloud DNS
Payments, consent, captcha, video, maps, reviews, forms, scheduling	Stripe, PayPal, OneTrust, Cookiebot, reCAPTCHA, Turnstile, YouTube, Vimeo, Trustpilot, Calendly, Typeform

Input

Field	Description
urls	Websites to analyze. Domains (example.com) or full URLs.
urlsText	Bulk paste: one website per line, or separated by commas. Header lines like website are ignored.
datasetId / datasetField	Read websites from a dataset, for example the output of another Actor. The field can be nested (company.website) and can hold a list.
includeDns	Detect email, DNS and SaaS providers from DNS records. On by default.
includeEvidence	Add the evidence list to every technology. On by default; turn it off for smaller output.
respectRobotsTxt	Skip sites whose robots.txt disallows the homepage (not charged). On by default.
dedupe	Analyze each domain only once. On by default.
maxItems	Stop after this many websites (0 = no limit).
maxConcurrency	Websites processed in parallel (max 50). Automatic by default: 20 at 1024 MB, 12 at 512 MB, 8 at 256 MB.
timeoutSecs	Timeout per website (default 20 seconds).
userAgent	Optional custom User-Agent.

Example:

{
    "urls": ["shopify.com", "wordpress.org", "https://vercel.com"],
    "includeDns": true,
    "includeEvidence": true
}

Output

One dataset row per website. The most useful fields are flattened for spreadsheets (cms, ecommerce, analytics, advertising, cdn, hosting, emailHosting, dnsProvider), and the full list is in technologies.

{
    "input": "wordpress.org",
    "domain": "wordpress.org",
    "url": "https://wordpress.org/",
    "status": "ok",
    "httpStatus": 200,
    "title": "Blog Tool, Publishing Platform, and CMS - WordPress.org",
    "technologies": [
        {
            "name": "WordPress",
            "categories": ["CMS", "Blog"],
            "version": "7.2",
            "confidence": "detected",
            "evidence": ["meta generator: WordPress 7.2-beta3-65005"]
        },
        {
            "name": "Nginx",
            "categories": ["Web server", "Reverse proxy"],
            "version": null,
            "confidence": "detected",
            "evidence": ["header server: nginx"]
        },
        {
            "name": "PHP",
            "categories": ["Programming language"],
            "version": null,
            "confidence": "implied",
            "evidence": ["implied by WordPress"]
        }
    ],
    "technologyCount": 11,
    "categories": { "CMS": ["WordPress"], "Web server": ["Nginx"] },
    "cms": ["WordPress"],
    "ecommerce": [],
    "analytics": ["Jetpack"],
    "cdn": [],
    "hosting": [],
    "emailHosting": [],
    "dnsProvider": [],
    "dns": {
        "mxHosts": ["..."],
        "nameservers": ["..."],
        "wwwCname": null,
        "ipv4": ["..."],
        "domainExists": true
    },
    "server": "nginx",
    "poweredBy": null,
    "redirects": [],
    "responseTimeMs": 310,
    "error": null,
    "checkedAt": "2026-10-29T15:00:00.000Z"
}

(Shortened. Real rows list every detected technology.)

confidence is detected when the technology itself left a trace, and implied when another detection implies it (for example WordPress implies PHP).

Status values

Status	Meaning	Charged
ok	The homepage loaded and was analyzed.	Yes
http_error	The site answered with an error (403, 404, 500...). Headers, cookies and DNS are still analyzed.	No
unreachable	Timeout, connection or TLS error, or the domain does not exist.	No
robots_disallowed	robots.txt does not allow fetching the homepage.	No
invalid_input	Not a valid public website.	No

The run also writes an OUTPUT record to the key-value store with counts per status and the most common technologies across all sites.

Pricing

Price change on Nov 16, 2026: the price per analyzed website goes from $0.002 to $0.0023 ($2.30 per 1,000). Failed sites stay free.

$2.00 per 1,000 analyzed websites ($0.002 per site), pay per event. Only websites with status ok are charged. Everything else is free. A small Actor start fee set by Apify applies per run.

If you set a maximum cost per run, the Actor stops cleanly when the limit is reached, and websites that were not analyzed are not charged.

How it works

1. Each input is cleaned into a domain or URL (invalid values, IP addresses and private hosts are rejected).
2. The homepage is fetched with redirects followed. If it fails, the www or non-www variant and the other scheme are tried.
3. HTTP headers, cookies, script and stylesheet URLs, meta tags and HTML patterns are matched against a hand-written fingerprint list.
4. In parallel, MX, TXT, NS, CNAME and A records are looked up and matched too.
5. Implied technologies are added and the result is grouped by category.

Only public information is used. The Actor does not log in, solve captchas, render JavaScript or collect personal data.

Limits

• Only the homepage (or the URL you give) is analyzed, and JavaScript is not executed. Technologies loaded only after user interaction, or only on inner pages, can be missed.
• Some sites block automated requests (often with a 403, mostly behind Cloudflare or similar bot protection). They are reported as http_error, their headers, cookies and DNS are still analyzed, and they are not charged. The Actor identifies itself honestly and does not try to get around these blocks.
• On random lists of domains (for example a sample of a top 1M list) expect roughly 65 to 75% ok: many such domains no longer resolve, are parked, or block datacenter traffic. On lists of real company websites the share is much higher.
• Long dashes in page titles and descriptions are replaced with a plain hyphen.
• Detections from DNS (for example SendGrid in an SPF record) show that the domain is set up for that service, not necessarily that it is actively used.

FAQ

Do I need proxies? No. One request per site from Apify's datacenter is enough for most websites.

How fast is it? Roughly 200 to 300 websites per minute at the default concurrency, depending on how fast the sites answer.

How much memory should I use? Keep the default 1024 MB for lists of hundreds or thousands of websites. 256 MB is enough for short lists. Memory does not change the price per website.

Can I feed it the results of another Actor? Yes. Use datasetId and point datasetField at the field that holds the website, for example website or company.domain.

A technology is missing or wrong. Open an issue on the Actor page with the website and the technology, and the fingerprint will be added or fixed.

Support

Questions or bugs: open an issue on the Actor's Issues tab. Contact: joesimo@mail.grokbot.com

More examples and the docs for all 6 of our Actors: github.com/Joe-Simo/apify-actors. Store page: https://apify.com/joesimo/tech-stack-detector