DevStudio's HAR Analyzer parses HTTP Archive files exported from Chrome DevTools, Firefox, Safari Web Inspector, or any tool that emits the standard HAR 1.2 schema, and turns the captured network log into a navigable inspection workspace inside your browser. Drop in one or several HAR files and the tool surfaces every request and response with full timing breakdowns — DNS lookup, TCP connect, TLS handshake, time to first byte, content download — plus headers, cookies, query strings, request and response bodies, and the resource initiator chain when the recording browser supplied it. Filter the request list by status code, MIME type, hostname, response time, or body content, and the tool dynamically updates summary statistics so you can see, for example, every 4xx and 5xx response on a particular API host or every request slower than two seconds. Side-by-side comparison of multiple HARs is built in, which is invaluable when you need to spot what changed between a working capture and a broken one — a vanished cookie, a new redirect chain, a slower handshake against a different IP. Common workflows include diagnosing a CORS preflight failure by inspecting the Access-Control headers, finding the request that caused a particular 502 response, recreating a curl command from a captured request, sharing a HAR with a backend engineer without giving them browser access, and triaging a slow page load by sorting requests by total time and zooming in on the timing breakdown of the slowest entry. Because the entire parser runs locally in your browser, the captured cookies, authorization headers, and request bodies — which often contain credentials, session tokens, or PII — never leave your device. The tool is safe for enterprise debugging where uploading a HAR to an online service would violate compliance, and it works offline once the page has loaded so you can analyze captures from an air-gapped environment.
A HAR file (HTTP Archive) is a JSON document that records every network request and response a browser made while a page loaded. It captures URLs, headers, cookies, request and response bodies, status codes, and detailed timing information for each phase of the request. Browsers including Chrome, Firefox, and Safari can export a HAR from their DevTools network panel, and many proxy and testing tools produce the same standard format.
In Chrome or Edge open DevTools, switch to the Network tab, ensure recording is on, reproduce the issue, then right-click the request list and choose Save all as HAR with content. Firefox offers the same option under Save All As HAR, and Safari Web Inspector exports from its Network timeline. Save the resulting JSON file and drop it into DevStudio HAR Analyzer to inspect every captured exchange offline.
HAR files frequently include sensitive information — Authorization headers, session cookies, request bodies with personal data — because they capture every header and body the browser sent and received. Before sharing a HAR with anyone outside your security boundary, sanitize it: redact bearer tokens, strip cookies, and remove request bodies that contain personal data. DevStudio HAR Analyzer runs locally so you can inspect a capture safely, but only share the cleaned version externally.
DevStudio HAR Analyzer parses your HAR file entirely in your browser, with no upload step and no telemetry on the contents. The parser, the timeline charts, the filters, and the side-by-side comparison all run client-side, so even a HAR containing production cookies or internal hostnames stays on your device. The tool also works offline once loaded, which makes it suitable for air-gapped or compliance-restricted environments.
Yes. DevStudio HAR Analyzer accepts up to several HAR captures at once and aligns their request lists side by side so you can quickly spot what differs between a working and broken recording. Common diagnostics include a missing redirect, a changed cookie, a new CORS preflight, or a slower TLS handshake against a different IP. The comparison highlights status, header, and timing differences without you having to scan two large logs by eye.