HAR Viewer
Open a HAR file — the HTTP Archive your browser's DevTools exports from the Network tab — and read it as a clean request table. See each request's method, status, resource type, transfer size, and timing, with a simple waterfall showing when each one happened. Headline metrics summarise total requests, bytes transferred, page load time, and how many responses were errors. Filter by type or URL to focus. Everything is parsed in your browser; the HAR never leaves your machine.
How to use the HAR Viewer
In your browser open DevTools, go to the Network tab, reproduce the activity you want to capture, then right-click any request and choose Save all as HAR (Chrome and Edge) or use the download/export button (Firefox, Safari). Paste the saved JSON here or load the .har file. The viewer lists every request with its status colour-coded — green for 2xx, grey for redirects, amber for 4xx, red for 5xx — alongside size and timing, and a waterfall bar positioned by when each request started.
Use the type dropdown to isolate scripts, stylesheets, images, fonts, documents, or XHR/fetch calls, and the search box to filter by URL. The summary cards give the at-a-glance picture: how many requests fired, how much was transferred, the total wall-clock span, the slowest single request, and the count of client and server errors. That's usually enough to spot the heavy image, the blocking script, or the failing API call behind a slow or broken page.
What a HAR file captures
HAR — HTTP Archive — is a JSON format that records the network activity of a page or session. Every browser's developer tools can export one, and the file contains an entry for each request: the full URL and method, request and response headers, status code, payload sizes, MIME type, and a detailed timing breakdown covering DNS, connect, TLS, send, wait, and receive phases. Because it's a complete, portable snapshot, a HAR is the standard artefact to attach to a bug report when a page is slow or a request misbehaves on someone else's machine.
Reading a raw HAR is unpleasant, though — it's deeply nested JSON, often megabytes in size, with one request's details buried under several levels of objects. A viewer turns it back into the waterfall you'd see in DevTools, but decoupled from the original session, so you can analyse a capture a colleague or customer sent you. The metrics that matter for performance work surface immediately: total transfer size tells you how heavy the page is, the load span tells you how long it took, and the per-request timing reveals which resource is the bottleneck. The error counts catch failed requests that might otherwise hide in a long list.
HAR files deserve care because they're thorough: they can include request and response headers, cookies, and sometimes posted form data or response bodies, which means they may carry authentication tokens or personal information. That's exactly why parsing one in the browser, with nothing uploaded, is the safe way to inspect a capture — and why you should still scrub or limit a HAR before sharing it. This viewer reads the structure for analysis and never transmits it.
Common use cases
- Performance triage. Spot the largest payloads and slowest requests dragging down a page load.
- Bug reports. Open a HAR a user attached to see exactly which request failed and with what status.
- Third-party audit. Filter to one domain to measure how much weight an external script or pixel adds.
- API debugging. Isolate XHR/fetch calls to check status codes and response sizes for a flow.