Browser Network & HAR Capture Inspector

Capture real Chromium network traffic from public webpages, export a HAR 1.2 file for every captured page, and get a compact diagnostic summary of XHR/fetch/API calls, redirects, failures, HTTP errors, slow resources, third-party domains, and transfer size.

This Actor is for developers who need a network diagnostic outcome, not a low-level CDP command runner or a broad endpoint-probing crawler.

Quick start

{
  "startUrls": [{ "url": "https://example.com" }],
  "captureDurationMs": 1500,
  "includeDetailedRows": false,
  "captureResponseBodies": false,
  "proxyConfiguration": { "useApifyProxy": false }
}

The default dataset receives one summary row per page. The HAR file is stored in the default key-value store under a key such as har-001-example.com.har.

Capture XHR/fetch only

{
  "startUrls": [{ "url": "https://googlechrome.github.io/samples/fetch-api/fetch-html.html" }],
  "captureDurationMs": 3000,
  "resourceTypes": ["xhr", "fetch"],
  "includeDetailedRows": true
}

What it returns

A page summary contains fields such as:

{
  "rowType": "summary",
  "passed": true,
  "inputUrl": "https://example.com/",
  "finalUrl": "https://example.com/",
  "requestCount": 7,
  "capturedRequestCount": 7,
  "xhrFetchCount": 2,
  "httpErrorCount": 0,
  "failedRequestCount": 0,
  "redirectCount": 1,
  "slowRequestCount": 1,
  "totalTransferBytes": 82431,
  "harKey": "har-001-example.com.har",
  "harUrl": "https://api.apify.com/v2/key-value-stores/.../records/har-001-example.com.har",
  "apiEndpoints": ["https://example.com/api/data"],
  "warnings": []
}

When includeDetailedRows is enabled, each captured request can additionally produce a compact network row with method, URL, resource type, status, initiator type when available, duration, transfer size, redirect/failure flags, and XHR/fetch/GraphQL/WebSocket classification.

HAR behavior

• HAR version: 1.2.
• One HAR artifact per processed page/session.
• Common auth, cookie, proxy-auth, and API-key headers are replaced with [REDACTED].
• Credential-like query parameter values are redacted in exported URLs/query strings.
• Response bodies are OFF by default.
• If body capture is explicitly enabled, only bounded textual bodies are embedded.
• maxHarBytes removes optional bodies first and then truncates recorded entries if required to keep the artifact within the configured limit.

Useful inputs

Input	Purpose	Default
startUrls	Public webpages to capture	required
waitUntil	Navigation completion condition	domcontentloaded
captureDurationMs	Extra time for late XHR/fetch	1500
includePatterns / excludePatterns	Glob filters for recorded requests	empty
resourceTypes	Optional resource-type filter	all
includeDetailedRows	Push one dataset row per request	false
captureResponseBodies	Embed bounded textual bodies in HAR	false
maxResponseBodyBytes	Per-response body cap	262144
maxRequestsPerPage	Recorded request cap	500
maxHarBytes	HAR artifact cap	8388608
maxPages	Page/session cap	up to 10 by default
slowThresholdMs	Slow-resource threshold	1000
proxyConfiguration	Apify/custom proxy	no proxy

When to use this Actor

• Export a HAR from a public JavaScript application without opening DevTools manually.
• Discover the XHR/fetch/API traffic a page generates at runtime.
• Find 4xx/5xx or failed browser requests.
• Identify third-party domains contacted by a landing page.
• Compare top slowest or largest resources.
• Run scheduled network diagnostics and save a HAR artifact for each run.

Safety and privacy

This Actor only supports public HTTP(S) destinations. It rejects localhost, loopback, RFC1918/private, link-local, metadata, multicast/reserved addresses, credential-bearing URLs, and non-web protocols. Browser redirects and subresources pass through the same validating egress proxy.

Authorization, Cookie, Set-Cookie, proxy credentials, and common API-key headers are redacted by default. Response bodies remain off unless you deliberately enable them. Do not put credentials or secrets into public task examples, README snippets, support tickets, or screenshots.

Blocked sites, 403/429 responses, robots/terms restrictions, CAPTCHAs, and access controls are target-specific limitations. This Actor is not positioned as a bypass tool.

Pricing

Launch pricing uses Pay Per Event:

• page-capture: $0.02 for each successful page/session capture
• standard apify-actor-start synthetic event remains at the platform-recommended default

Examples before the Actor-start event:

• 1 successful page capture: $0.02
• 5 successful page captures: $0.10
• 10 successful page captures: $0.20

Security-blocked/preflight-rejected pages are not charged as successful captures. Use Apify's Maximum cost per run to bound PPE spend.

Which ProdukDigitalAli browser tool should I choose?

Tool	Choose it when...
Puppeteer Browser Automation Runner	You want simple Chrome/Puppeteer-style deterministic workflows.
Selenium Browser Automation Runner	You think in WebDriver locators or maintain Selenium knowledge.
Chrome DevTools Automation Runner	You need low-level CDP commands, emulation, rendering, or diagnostics.
Playwright Browser Automation Runner	You want deterministic JSON navigation/click/form/extraction/assertion workflows.
Browser Network & HAR Capture Inspector	You need HAR export, XHR/fetch/API visibility, failed requests, timing, and third-party domains.

Limits

• Chromium only in the MVP.
• Up to 20 input URLs, with bounded maxPages.
• Detailed request rows are opt-in to keep dataset size and cost predictable.
• WebSocket support is diagnostic/count-oriented in the MVP; this is not a packet/frame recorder.
• GraphQL classification is heuristic (URL/operation indicators), not schema introspection.

Support

When reporting an issue, include the Actor run ID, sanitized input, target URL if it is public, and the relevant errorCode. Never post API keys, cookies, bearer tokens, passwords, proxy credentials, or private HAR files containing sensitive data.