speedlify2

Lighthouse lab metrics, Axe, and Chrome UX Report field data for a list of sites recorded and compared over time. Follow us on Mastodon or Bluesky.

Charts Compare Measurement runs Embed a score Open to Work

mistral.ai

https://mistral.ai/

59PERF 95A11Y 96BEST 100SEO ✗AXE ✗CWV

Ranks Overall #1,735 Developer Tooling #84 AI #8

Data from Oct 16, 2026, 11:06 AM UTC old Measured every 2 days · 21 runs since Aug 31, 2026

Elsewhere: PageSpeed Insights · Calibre · Treo · CrUX Vis · Tame the Bots · Scoring Calculator · Wayback Machine

Lab data is measured using Lighthouse, Mobile mode unless otherwise specified. Field data is reported from Chrome UX Report (using p75 measurements).

Stack

Astro Built with Astro
Uses ReactReact
Netlify Hosted by Netlify

Filmstrip LAB Throttled

How the page drew during the measured load, at 8 points across 17.61 s. Compare this filmstrip against another site.

  1. 1.57 s
  2. 4.40 s
  3. 6.61 s
  4. 8.81 s
  5. 11.01 s
  6. 13.21 s
  7. 15.41 s
  8. 17.61 s

Lighthouse Scores LAB Simulated

59
Performance
↑ +2 (+3.5%)
95
Accessibility
no change
96
Best Practices
no change
100
SEO
↑ +8 (+8.7%)

Metrics LAB FIELD

Core Web Vitals: failing 3 of 3 assessed from real Chrome users.

Field figures are the p75 across real Chrome users over a 28-day trailing window (2026-09-17 → 2026-10-14), url-level.

Metric Lab Simulated Field
Value Change / range Trend p75 Distribution
LCP CWV Largest Contentful Paint 13.67 s poor ↑ +475 ms (+3.6%) 12.43 s – 19.37 s 4.96 s poor 31.6% 35.9% 32.6%
INP CWV Interaction to Next Paint — field only, no lab run can measure it n/a — 165 ms good 81.1% 14.4% 4.5%
CLS CWV Cumulative Layout Shift 0.012 good ↑ +0.001 (+9.1%) 0.006 – 0.017 0.070 good 77.9% 6% 16.1%
TBT Total Blocking Time 292 ms needs work ↓ -92 ms (-24%) 236 ms – 772 ms n/a —
FCP First Contentful Paint 3.60 s poor ↑ +150 ms (+4.4%) 1.69 s – 5.68 s 1.06 s good 89.3% 6.1% 4.7%
Speed Index 4.51 s needs work ↑ +473 ms (+11.7%) 4.04 s – 6.18 s n/a —
TTFB Time to First Byte 522 ms good ↓ -151 ms (-22.4%) significant 472 ms – 673 ms 537 ms good 85.8% 9.9% 4.3%

LCP Subparts LAB Throttled

The subparts attribute LCP to your server, resource discovery, download, or rendering. They come from one throttled load, the same one as the filmstrip, so they add up to that load’s LCP rather than the simulated Lab LCP above.

element render delay 16.65 s
time to first byte 634 ms

LCP LAB Simulated vs LCP FIELD

Lab 13.67 s
Field p75 4.96 s

The lab LCP is 2.76× the field p75 — real users likely have above-average hardware or network connectivity.

Page weight

Metric Value Change / range Trend
Page Weight 2.45 MB ↑ +214.7 kB (+9.9%) significant 2.17 MB – 3.52 MB
Requests 69 ↑ +5 (+7.8%) significant 63 – 88
Third-party Weight 793 kB ↑ +177.0 kB (+29.6%) significant 595 kB – 793 kB
Third-party CPU 179 ms ↓ -58 ms (-24.4%) significant 178 ms – 239 ms
Unused JS 445 kB ↓ -7.9 kB (-1.8%) 435 kB – 458 kB
Unused CSS 0 B → no change 0 B – 0 B

By resource type

Type Transfer Share Size Requests
script 40.4% 985 kB 10
image 30.4% 741 kB 33
font 24.9% 608 kB 8
document 2.7% 66.3 kB 1
stylesheet 1.5% 37.4 kB 3
other 0.2% 3.8 kB 3
Total 2.45 MB 69

Third parties 3 entities

Entity Transfer Size Main thread Simulated
axept.io scriptfont 451 kB 86 ms
Google Tag Manager script 342 kB 93 ms
Google Analytics 544 B 0 ms
Total 793 kB 179 ms

Without JavaScript

The same load, photographed twice at 412×823: once with scripts disabled, once as rendered.

Scripts disabled
As rendered

11.4% of pixels differ — parts of the page change. A shifted layout counts as difference, so a <noscript> banner alone can account for a large figure.

Page health

Metric Value Change / range Trend
Main Thread Work 4.15 s ↓ -1077 ms (-20.6%) significant 4.15 s – 8.19 s
Long Tasks 11 ↑ +1 (+10%) 10 – 13
DOM Elements 2,575 ↓ -151 (-5.5%) significant 2,575 – 3,688
Axe Violations Violating nodes from a full Axe CLI run (more than Lighthouse reports) 6 → no change 5 – 12
A11y Failures 2 → no change 2 – 3
A11y Failing Nodes 7 ↑ +1 (+16.7%) significant 5 – 11

Main thread work

Task Time Share Duration Simulated
other 44% 1.83 s
script evaluation 27% 1.12 s
style layout 14% 580 ms
paint composite render 9% 393 ms
script parse compile 4% 150 ms
parse HTML 1% 53 ms
garbage collection 1% 23 ms
Total 4.15 s

Axe violations axe-core 4.13.0, full rule set

6 violating nodes across 3 rules, 2096 passing. This is a standalone Axe CLI run and is stricter than the subset Lighthouse already includes.

  • critical button-name — Buttons must have discernible text 4 nodes
    Offending elements
    1. .p-2.z-5.lg\:hidden > .lg\:gap-4.p-2[href$="mistral-large-4/"] > .h-7.min-w-7.bg-action-ghost-1 
      <button class="select-none flex relative items-center justify-center disabled:text-text-muted overflow-hidden disabled:pointer-events-none h-7 min-w-7 bg-action-ghost-1 hover:bg-action-ghost-2 text-t…
    2. .outline 
      <button class="js-theme-option cursor-pointer shrink-0 flex items-center justify-center size-7 rounded hover:bg-surface-brand-primary bg-surface-brand-primary outline outline-border-primary" data-the…
    3. button[data-theme="light"] 
      <button class="js-theme-option cursor-pointer shrink-0 flex items-center justify-center size-7 rounded hover:bg-surface-brand-primary" data-theme="light">
    4. button[data-theme="dark"] 
      <button class="js-theme-option cursor-pointer shrink-0 flex items-center justify-center size-7 rounded hover:bg-surface-brand-primary" data-theme="dark">
  • moderate landmark-unique — Landmarks should have a unique role or role/label/title (i.e. accessible name) combination 1 node
    Offending element
    1. .md\:justify-start 
      <nav class="flex justify-center md:justify-start items-center p-4 md:p-0 gap-6 w-full border-b border-border-primary md:border-0">
  • moderate region — All page content should be contained by landmarks 1 node
    Offending element
    1. .h-64 
      <div class="w-full h-64 relative overflow-hidden">

Accessibility failures 2 of 30 audits

  • button-name — Buttons do not have an accessible name 4 nodes
  • label-content-name-mismatch — Elements with visible text labels do not have matching accessible names. 3 nodes

Best Practices failures 2 of 19 audits

  • image-size-responsive — Serves images with low resolution 2 items
  • valid-source-maps — Missing source maps for large first-party JavaScript 5 items

Hygiene

HTTPS HSTS CSP Clickjacking protection Origin isolation Trusted Types bfcache h3

Measurement log

Field history 25 weeks

Weekly p75 from the CrUX History API, backfilled in a single request.

Week ending LCPINPCLSFCPTTFB
2026-10-10 5.03 s 170 ms 0.070 1.07 s 544 ms
2026-10-03 5.16 s 181 ms 0.070 1.09 s 562 ms
2026-09-26 5.06 s 183 ms 0.080 1.09 s 562 ms
2026-09-19 4.96 s 184 ms 0.120 1.10 s 563 ms
2026-09-12 4.70 s 179 ms 0.180 1.08 s 558 ms
2026-09-05 4.18 s 168 ms 0.230 1.03 s 540 ms
2026-08-29 3.98 s 155 ms 0.240 987 ms 529 ms
2026-08-22 3.74 s 145 ms 0.160 950 ms 513 ms
2026-08-15 3.60 s 136 ms 0.050 919 ms 493 ms
2026-08-08 3.60 s 133 ms 0.020 932 ms 486 ms
2026-08-01 3.52 s 130 ms 0.010 913 ms 469 ms
2026-07-25 3.39 s 126 ms 0.010 884 ms 458 ms

Runs

Every stored run for this URL, newest first.

When Scores PerfLCPCLSTBTWeight RunsSpreadBenchTook Archive
Oct 16, 2026, 11:06 AM UTC 59 13.67 s 0.012 292 ms 2.45 MB 3/3 ±14 3,099 +25% 97.8s Wayback
Oct 14, 2026, 6:02 AM UTC 57 13.20 s 0.011 384 ms 2.23 MB 3/3 ±8 2,469 -1% 104.6s Wayback
Oct 9, 2026, 3:10 AM UTC 51 13.29 s 0.017 399 ms 2.19 MB 3/3 ±6 2,473 -1% 105.1s Wayback
Oct 7, 2026, 2:21 AM UTC 53 12.43 s 0.010 337 ms 2.17 MB 3/3 ±3 2,475 -1% 105.1s Wayback
Oct 4, 2026, 4:38 AM UTC 55 18.29 s 0.006 376 ms 3.43 MB 3/3 ±4 3,336 +34% 104.3s Wayback
Sep 30, 2026, 3:53 PM UTC 63 18.19 s 0.007 304 ms 3.43 MB 3/3 ±8 2,953 +19% 114.7s Wayback
Sep 28, 2026, 10:53 AM UTC 59 19.37 s 0.007 398 ms 3.43 MB 3/3 ±1 2,650 +6% 116s Wayback
Sep 25, 2026, 11:26 AM UTC 54 17.78 s 0.006 687 ms 3.44 MB 3/3 ±4 2,373 -5% 110.7s Wayback
Sep 23, 2026, 5:42 AM UTC 58 19.02 s 0.007 443 ms 3.43 MB 3/3 ±1 2,497 0% 117.1s Wayback
Sep 21, 2026, 1:14 AM UTC 60 18.82 s 0.007 459 ms 3.52 MB 3/3 ±2 2,454 -1% 112.1s Wayback
Sep 18, 2026, 1:47 PM UTC 64 18.25 s 0.006 293 ms 3.44 MB 3/3 ±11 2,965 +19% 108.6s Wayback
Sep 16, 2026, 3:59 AM UTC 59 18.45 s 0.006 426 ms 3.42 MB 3/3 ±1 2,482 0% 116.8s Wayback
Sep 14, 2026, 1:10 AM UTC 66 18.89 s 0.008 236 ms 3.35 MB 3/3 ±11 3,331 +34% 104.9s Wayback
Sep 11, 2026, 11:04 PM UTC 54 17.58 s 0.006 715 ms 3.34 MB 3/3 ±1 2,351 -6% 111.5s Wayback
Sep 9, 2026, 5:53 PM UTC 58 17.28 s 0.007 518 ms 3.22 MB 3/3 ±4 2,485 0% 63.2s Wayback
Sep 7, 2026, 3:13 PM UTC 56 18.40 s 0.006 525 ms 3.30 MB 3/3 ±2 1,872 -25% 60s Wayback
Sep 5, 2026, 12:32 PM UTC 53 17.07 s 0.007 772 ms 3.21 MB 3/3 ±3 2,159 -13% 59.5s Wayback
Sep 4, 2026, 1:15 AM UTC 58 17.66 s 0.007 249 ms 3.30 MB 3/3 ±9 3,418 +37% 53.6s Wayback
Sep 2, 2026, 11:22 PM UTC 64 17.50 s 0.008 329 ms 3.30 MB 3/3 ±2 2,855 +15% 56.2s Wayback
Sep 1, 2026, 10:48 PM UTC 56 18.02 s 0.007 488 ms 3.38 MB 3/3 ±11 2,386 -4% 58.3s Wayback
Aug 31, 2026, 9:06 PM UTC 60 17.52 s 0.007 405 ms 3.29 MB 3/3 ±3 2,492 0% 60.5s Wayback