Website Tech Stack Detector

Give it a list of domains or URLs. For each one it fetches the homepage once and tells you what the site runs on: CMS, ecommerce platform, analytics, tag manager, CDN, web server, hosting, frameworks, payment processors and the email provider, from a rule set of more than 6,700 technologies. Every record shows the evidence for each detection and links to the page it came from.

Use cases

• Sales and lead qualification. Find every company in a list that runs Shopify, WordPress or a given analytics tool, then sort by what they use.
• Competitive and market research. See which platforms a set of competitor or customer sites are built on.
• Agency and migration prospecting. Spot outdated versions, sites without a CDN, or stores on a platform you replace.
• Data enrichment. Add a technology column to a spreadsheet of domains. Flat fields (cms, ecommerce, analytics, cdn, email_provider and others) drop straight into a CSV.
• AI agents. An agent that is asked "what does this company use?" can call this with a domain and get a structured, checkable answer in about a second, with no browser to drive.

What it does not do

• It does not run a browser or JavaScript. Technologies that can only be seen by executing JavaScript in a page (for example a library found only as a global variable) are not detected, and a site that builds its whole page with JavaScript shows only what is in its response headers, its static markup and its DNS.
• It reads one page per domain, the one you give it. It does not crawl, log in, or get around bot protection. If a site answers with an error or a challenge page, that response is analyzed as it is, fetch_status says http_error, and the result may be incomplete.
• It does not collect or output emails, phone numbers, names or any page text. Page content is read only to match technology signatures and is never copied into the output.
• Detection is evidence-based, not certain. A confidence score is given for each technology and the evidence type is listed so you can judge it. A technology that does not leave a trace in the page, headers or DNS cannot be seen.
• DNS TXT records show that a domain verified itself with a service (for example a mail or collaboration tool). Those appear in technologies with evidence: ["dns:TXT"] and are kept out of the flat fields.

Sample output

Input:

{ "domains": ["nginx.org"] }

One record (shortened to two technologies):

{
  "input": "nginx.org",
  "domain": "nginx.org",
  "final_url": "https://nginx.org/",
  "status": 200,
  "fetch_status": "ok",
  "fetch_note": "",
  "technologies": [
    { "name": "Nginx", "categories": ["Web servers", "Reverse proxies"], "version": "1.31.3",
      "confidence": 100, "evidence": ["headers"], "website": "https://nginx.org/en" },
    { "name": "HSTS", "categories": ["Security"], "version": null,
      "confidence": 100, "evidence": ["headers"], "website": "https://www.rfc-editor.org/rfc/rfc6797#section-6.1" }
  ],
  "technology_count": 2,
  "cms": [], "ecommerce": [], "analytics": [], "tag_managers": [], "cdn": [],
  "web_server": ["Nginx"], "hosting": [], "email_provider": [],
  "javascript_frameworks": [], "web_frameworks": [], "programming_languages": [],
  "payment_processors": [], "mx_hosts": [],
  "fetched_at": "2026-10-18T09:00:00Z",
  "verify_url": "https://nginx.org/",
  "fingerprints": "692e0ec"
}

How to check it yourself. Open the verify_url, or run curl -I https://nginx.org/ and read the Server header: it prints nginx and a version number, which is the evidence behind the Nginx entry and its version. The nginx documentation says nginx puts its version in that header. For a Shopify store, view the page source and search for cdn.shopify.com. For email, look up the domain's MX records with any DNS tool and compare to mx_hosts.

A site that cannot be reached returns a clean record and is not charged:

{ "input": "nonexistent-domain-abc123xyz987.com", "domain": "nonexistent-domain-abc123xyz987.com",
  "final_url": null, "status": null, "fetch_status": "unreachable",
  "fetch_note": "the domain name does not resolve", "technologies": [], "technology_count": 0,
  "verify_url": "https://nonexistent-domain-abc123xyz987.com/" }

(The real record also carries the empty flat fields shown above.)

Configuration

Input	Type	Required	Default	Description
domains	list of text	yes	["wordpress.org"]	Domains or page URLs. A bare domain is fetched over https, and over http if https is not available. Up to 1000 per run. Repeats are fetched once.
minConfidence	integer 0 to 100	no	50	Only report technologies at or above this confidence. Weak single signals score under 50. Use 0 to see everything.
includeDns	true or false	no	true	Look up MX, TXT and NS records. This is how the email provider is found. Turn off for the fastest run.

Only the default web ports (80 and 443) are fetched. IP addresses, internal hostnames, URLs with a username or password, and non-web schemes are refused with a message, and are not charged.

Output fields

Field	Meaning
input, domain	What you gave, and the domain in ASCII form with a leading www. removed
final_url, status	The page that answered after redirects, and its HTTP status (null if nothing answered)
fetch_status	ok, http_error (the site answered with an error status, analyzed anyway), not_html, or unreachable
fetch_note	Plain-language note when something limited the result
technologies	List of name, categories, version (if found), confidence (0 to 100), evidence (url, html, text, css, scripts, scriptSrc, meta, headers, cookies, dom, dns:MX, dns:TXT, dns:NS, implied) and website
technology_count	Length of technologies
cms, ecommerce, analytics, tag_managers, cdn, web_server, hosting, email_provider, javascript_frameworks, web_frameworks, programming_languages, payment_processors	Flat lists of technology names, for spreadsheets. email_provider comes from MX records only
mx_hosts	Up to 5 mail server hostnames, lowest preference first
fetched_at	UTC time of the fetch
verify_url	The page that was fetched, or attempted if it failed
fingerprints	Short id of the rule-set version used

What can go wrong

fetch_status or message	What it means	What to do
unreachable, "the domain name does not resolve"	The name has no address	Check the spelling
unreachable, "did not answer within 15 seconds"	The site is down or blocking	Try again later
http_error with status 403 or 429	The site refused the request or showed a bot check	Results come from headers and DNS only. Do not rely on a short list
not_html	The address is a file, not a page	Give the page URL
"certificate could not be verified" in fetch_note	The page was read anyway; its TLS certificate is invalid	Treat the site as misconfigured
Entry skipped in the run log	The entry was not a valid public domain or URL	Fix it. It is not charged

Pricing

Pay per event: $0.0015 per domain analyzed, about $1.50 per 1,000 domains, plus Apify's small start fee per run. A domain is charged only if the site answered over HTTP, including an error page that was still analyzed. Unreachable sites, duplicates and invalid entries are free.

Where the rules come from

Detection uses the community-maintained Wappalyzer rule set, enthec/webappanalyzer, licensed GPL-3.0, vendored at commit 692e0ec (2026-10-16), with a few documented additions. See SOURCES.md for the license, the list of what is and is not included, and where each test value came from.