Caliper

A real report

Quill's “generate a social post” button

This is the exact HTML file Caliper writes — unchanged, except that Quill's backend address is masked.

A summary of a real Caliper report, then the exact file it wrote.

Nov 14, 2026caliper-ai 0.6.0--runs 2 --server-runs 1 --stopOpen the file on its own →

Summary of the real report

1514 ms

click → first word, the one warm run

  • click → request sent3 ms within its 200 ms limit
  • browser: response headers received926 ms over its 600 ms limit
  • headers → first paint585 ms over its 100 ms limit

Verdict

People waited 1514 ms for the first word in the one warm run.

  • In the browser: headers → first paint — 5.8× its limit, in 1 of 1 warm runs.
  • Inside the server: postgres — 5.1× its limit, in 1 of 1 marked visits.

Advice · 2 things to look at

  • The first run (a fresh browser) waited 132 ms longer for the first word than the one warm run (1646 vs 1514 ms).

    why

    CORS preflight: 183 ms on the first run, not in the warm runs (+183 ms, more than its 100 ms limit). Caliper can't tell why from timings alone.

  • article is 55% of the prompt (3207 of 5827 bytes, labelled user).

    why

    Caliper's own threshold: one named part is 50% or more of the prompt's measured bytes, more than all the rest together. These are bytes your app reported, not tokens; Caliper didn't measure what trimming it would change.

Inside the server

postgres 506 ms

5.1× its 100 ms limit, in 1 of 1 marked visits; 48% of the time to the server's first bytes sent.

The unchanged file Caliper wrote, every number and detail.

This is the exact HTML file Caliper writes — unchanged, except that Quill's backend address is masked.