A
Before
BASELINENo capture selected
YOUR CAPTURES. YOUR BROWSER.
What changed
on the wire?
Two network captures. One clear comparison.
Find the new errors, heavier responses, and extra requests.
01 Add your captures
B
After
COMPARISONNo capture selected
Files stay in this browser. No uploads or accounts.
02 The difference
SYNTHETIC DEMOChanges worth a look
| Request group | Changes | Calls A → B | Body bytes change | Median change | Errors A → B |
|---|
Take the findings with you.
Copy a report into your issue, PR, or debugging notes.
Exports use request IDs by default. Headers, cookies, and bodies are excluded.
How to read this comparison
Requests are grouped by method and normalized URL. Repeated requests are compared as a group, using counts, body-byte totals, HTTP error counts, and median duration. Query parameters are included in matching unless you explicitly ignore them above. Their values are hidden in displayed URLs.
Missing or negative HAR measurements are unknown, not zero. Coverage is shown alongside the metrics. Response body bytes are not total network transfer or page weight.
Two captures show observations, not proof of a performance regression or its cause. Cache state, network conditions, device load, and user actions can change the result. Capture the same flow under similar conditions and repeat before drawing conclusions.
01 / CAPTURE
Record the same flow.
Open your browser’s Network panel, record a session, and export it as HAR. Repeat after your change.
02 / COMPARE
See the changes together.
Drop both captures here. Review added requests, HTTP errors, body sizes, and observed durations.
03 / SHARE
Bring evidence to the issue.
Export a compact report with request IDs. Choose whether to include URL paths when sharing.