Performance
· 16 checks — Lighthouse + waterfall + sustainability rolled into one auditable list.Performance-focused optimization plan with CWV assessment, top offenders, and time estimates
First Contentful Paint First Contentful Paint — how long until the browser renders the first piece of content. Under 1.8s is good.
4.48 s
Largest Contentful Paint Largest Contentful Paint — how long until the largest visible element loads. Under 2.5s is good.
22.56 s
Total Blocking Time Total Blocking Time — total time the main thread was blocked, preventing user input. Under 200ms is good.
3.92 s
Cumulative Layout Shift Cumulative Layout Shift — measures visual stability. How much the page layout shifts during loading. Under 0.1 is good.
0.000
Speed Index Speed Index — how quickly content is visually displayed during load. Under 3.4s is good.
13.31 s
Time to Interactive Time to Interactive — how long until the page is fully interactive and responds to user input. Under 3.8s is good.
26.62 s
Page Load Progression
FPage Weight BudgetAction4.7 MB transferred, 145 requestsFIX
Convert images to WebP/AVIF and resize to display dimensions to reduce transfer size.
Image bandwidth is high — modern formats (WebP/AVIF) and resizing typically cut it 50%+.
Source: web.dev
Large JavaScript bundles delay interactivity. Split code by route or defer non-critical scripts.
JavaScript bundle is large — code-split routes and lazy-load off-screen components to defer.
Source: web.dev
Subset fonts to include only used characters, or switch to system font stacks.
Font payload is large — subset to used glyphs only, or fall back to system fonts (zero load time).
Source: web.dev font loading
Each request adds latency. Bundle small files, use sprites, or eliminate unnecessary requests.
High HTTP request count — bundling, sprite-ing, and HTTP/2 server push can reduce per-request overhead.
Source: web.dev
DThird-Party ImpactAction100% third-party, 0 ms blockingFIX
FJS Execution CostAction16731ms total JS executionFIX
Main Thread Breakdown
| Script | Total | Scripting | Parse/Compile | Party |
|---|---|---|---|---|
| https://cohere.com/_next/static/chunks/35a9f203-8f15a8e7798acbe6.js | 6287ms | 5937ms | 58ms | 3rd |
| https://cohere.com/_next/static/chunks/3372-d43a11ed3cf32462.js | 3031ms | 2616ms | 35ms | 3rd |
| Unattributable | 1582ms | 81ms | 0ms | 3rd |
| https://cohere.com/ | 1530ms | 49ms | 112ms | 3rd |
| https://cohere.com/_next/static/chunks/6815-980a790505f7a80c.js | 898ms | 96ms | 13ms | 3rd |
| https://www.googletagmanager.com/gtm.js?id=GTM-5W6STPC | 520ms | 415ms | 87ms | 3rd |
| https://cdn-cookieyes.com/client_data/72267d03a3814ff1204bd164/banner.js | 508ms | 244ms | 4ms | 3rd |
| https://cohere.com/_next/static/chunks/2d1400c4.33880f77071ba18f.js | 390ms | 306ms | 57ms | 3rd |
| https://t.contentsquare.net/uxa/9298be563ec16.js | 383ms | 190ms | 71ms | 3rd |
| https://www.googletagmanager.com/gtag/js?id=G-S33PXJMLTN&cx=c>m=4e6460h1 | 359ms | 271ms | 85ms | 3rd |
Showing top 10 of 21 scripts
FJS BundlesAction61 scripts, 918 KB unusedFIX
Downloaded but never executed on this page.
| Script | Size | Unused | Minified | Party |
|---|---|---|---|---|
| https://cohere.com/_next/static/chunks/8043-a2ddd12a50120d22.js | 197 KB | 115 KB (59%) | ✓ | 3rd |
| https://cohere.com/_next/static/chunks/35a9f203-8f15a8e7798acbe6.js | 120 KB | 114 KB (95%) | ✓ | 3rd |
| https://t.contentsquare.net/uxa/9298be563ec16.js | 127 KB | 99 KB (78%) | ✓ | 3rd |
| https://www.googletagmanager.com/gtm.js?id=GTM-5W6STPC | 167 KB | 81 KB (49%) | ✓ | 3rd |
| https://cohere.com/_next/static/chunks/79a5c8c7-8d2e1a2c5f91247c.js | 85 KB | 76 KB (90%) | ✓ | 3rd |
| https://www.googletagmanager.com/gtag/js?id=G-S33PXJMLTN&cx=c>m=4e6460h1 | 169 KB | 69 KB (41%) | ✓ | 3rd |
| https://cohere.com/_next/static/chunks/2d1400c4.33880f77071ba18f.js | 78 KB | 46 KB (59%) | ✓ | 3rd |
| https://cohere.com/_next/static/chunks/9979-24400beca6c34fe1.js | 49 KB | 42 KB (86%) | ✓ | 3rd |
| https://cohere.com/_next/static/chunks/4069-0a5c37214518f62c.js | 57 KB | 41 KB (72%) | ✓ | 3rd |
| https://cohere.com/_next/static/chunks/56ad31fa-0b649efe885326be.js | 76 KB | 41 KB (53%) | ✓ | 3rd |
Consider code splitting or tree shaking to reduce unused code.
This bundle has high unused code — code-split or tree-shake to ship only what executes.
Source: web.dev / Lighthouse coverage
Consider code splitting or tree shaking to reduce unused code.
This bundle has high unused code — code-split or tree-shake to ship only what executes.
Source: web.dev / Lighthouse coverage
Consider code splitting or tree shaking to reduce unused code.
This bundle has high unused code — code-split or tree-shake to ship only what executes.
Source: web.dev / Lighthouse coverage
Consider code splitting or tree shaking to reduce unused code.
This bundle has high unused code — code-split or tree-shake to ship only what executes.
Source: web.dev / Lighthouse coverage
Consider code splitting or tree shaking to reduce unused code.
This bundle has high unused code — code-split or tree-shake to ship only what executes.
Source: web.dev / Lighthouse coverage
BImage Optimization53 images, 0 KB saveableREVIEW
CGreen HostingActionWhether the site is served from green-energy infrastructureREVIEW
Green Hosting
No green hosting detected
BHTTP Cachingpublic, max-age=0, must-revalidateREVIEW
public, max-age=0, must-revalidate
| Directive | Value | Meaning |
|---|---|---|
| public | — | Any cache may store this response |
| max-age | 0 | Cache for 0 seconds |
| must-revalidate | — | Must recheck with server after expiry |
A+Text CompressionAll text resources are compressedPASS
A+Font Loading6 fonts (424 KB)PASS
Web fonts
6
424 KB total
Render-blocking
0
of 6
Dominant font-display
swap
Most common across fonts
Font loading timeline
- Subset this font — over 100 KB suggests Latin Extended or full glyph coverage that most pages don't need
- Convert to woff2 for ~30% smaller file size
- Subset this font — over 100 KB suggests Latin Extended or full glyph coverage that most pages don't need
- Subset this font — over 100 KB suggests Latin Extended or full glyph coverage that most pages don't need
Optimization checklist
- Preload critical fonts (priority=high)
- Use woff2 format for all fonts
- Set font-display to swap, optional, or fallback
- Subset large fonts (≤100 KB each)
Fixing the unchecked items could save ~179 KB and ~358 ms
A+Resource CachingAll resources properly cachedPASS
A+Critical Rendering PathNo render-blocking resourcesPASS
A+Resource Hints20 hints, 0 missing preconnectsPASS
Current Resource Hints
20 resource hints configured
A+Render-Blocking ResourcesNo render-blocking resources detectedPASS
A+Third-Party ResourcesNo third-party resources detectedPASS
Network Waterfall145 requests over 0msINFO
Performance analysis not available for this result.
Third-Party Script CostPer-script blocking time, transfer cost, and cache headersINFO
| Script | Category | Execution | Transfer | Unused | Monthly Cost | Verdict |
|---|---|---|---|---|---|---|
cohere.com cohere.com | Other | 6287ms | 120 KB | 95% | €101/mo | Costly |
cohere.com cohere.com | Other | 3031ms | 51 KB | 46% | €48/mo | Costly |
Unattributable Unattributable | Other | 1582ms | 0 KB | — | €25/mo | Costly |
cohere.com cohere.com | Other | 1530ms | 69 KB | — | €24/mo | Costly |
cohere.com cohere.com | Other | 898ms | 26 KB | 84% | €14/mo | Costly |
Google Tag Manager www.googletagmanager.com | Tag Manager | 520ms | 167 KB | 49% | €8/mo | Costly |
cdn-cookieyes.com cdn-cookieyes.com | Other | 508ms | 8 KB | — | €8/mo | Costly |
cohere.com cohere.com | Other | 390ms | 78 KB | 59% | €6/mo | Costly |
t.contentsquare.net t.contentsquare.net | Other | 383ms | 127 KB | 78% | €6/mo | Costly |
Google Tag Manager www.googletagmanager.com | Tag Manager | 359ms | 169 KB | 41% | €6/mo | Costly |
cohere.com cohere.com | Other | 311ms | 54 KB | — | €5/mo | Costly |
cohere.com cohere.com | Other | 307ms | 21 KB | — | €5/mo | Costly |
cohere.com cohere.com | Other | 118ms | 197 KB | 59% | €2/mo | Optional |
cohere.com cohere.com | Other | 100ms | 158 KB | — | €2/mo | Optional |
cohere.com cohere.com | Other | 68ms | 6 KB | — | €1/mo | Optional |
cohere.com cohere.com | Other | 68ms | 76 KB | 53% | €1/mo | Optional |
cohere.com cohere.com | Other | 57ms | 28 KB | 94% | €1/mo | Optional |
cdn-cookieyes.com cdn-cookieyes.com | Other | 54ms | 9 KB | — | €1/mo | Optional |
cohere.com cohere.com | Other | 54ms | 85 KB | 90% | €1/mo | Optional |
cohere.com cohere.com | Other | 54ms | 26 KB | 83% | €1/mo | Optional |
cohere.com cohere.com | Other | 50ms | 59 KB | 52% | €1/mo | Optional |
These scripts may cost more than they're worth
- cohere.com adds 6287ms and costs ~€101/month
- cohere.com adds 3031ms and costs ~€48/month
- Unattributable adds 1582ms and costs ~€25/month
- cohere.com adds 1530ms and costs ~€24/month
- cohere.com adds 898ms and costs ~€14/month
- Google Tag Manager adds 520ms and costs ~€8/month
- cdn-cookieyes.com adds 508ms and costs ~€8/month
- cohere.com adds 390ms and costs ~€6/month
- t.contentsquare.net adds 383ms and costs ~€6/month
- Google Tag Manager adds 359ms and costs ~€6/month
- cohere.com adds 311ms and costs ~€5/month
- cohere.com adds 307ms and costs ~€5/month
100% of JavaScript execution time is spent on third-party scripts. Consider auditing which scripts are essential.
When third-party JS execution time exceeds your own, performance gains from frontend work are capped by code you don't own.
Learn more ▾ ▴
Every millisecond of third-party JS competes with your own for main-thread time. If their share is bigger than yours, optimization on your code base barely moves the needle. Audit, defer, or remove third-parties before further frontend optimization. Use Lighthouse's third-party audit to identify the worst offenders.
Source: web.dev
cohere.com takes 6287ms of CPU time. Consider loading it asynchronously or replacing it with a lighter alternative.
This script has high main-thread execution time — optimize hot paths or defer.
Source: web.dev
95% of cohere.com's code is unused. The script may be loading features you don't use.
Bundle has high unused-code ratio — tree-shaking and route-splitting recover the wasted bytes.
Source: web.dev
cohere.com takes 3031ms of CPU time. Consider loading it asynchronously or replacing it with a lighter alternative.
This script has high main-thread execution time — optimize hot paths or defer.
Source: web.dev
Unattributable takes 1582ms of CPU time. Consider loading it asynchronously or replacing it with a lighter alternative.
This script has high main-thread execution time — optimize hot paths or defer.
Source: web.dev
cohere.com takes 1530ms of CPU time. Consider loading it asynchronously or replacing it with a lighter alternative.
This script has high main-thread execution time — optimize hot paths or defer.
Source: web.dev
cohere.com takes 898ms of CPU time. Consider loading it asynchronously or replacing it with a lighter alternative.
This script has high main-thread execution time — optimize hot paths or defer.
Source: web.dev
84% of cohere.com's code is unused. The script may be loading features you don't use.
Bundle has high unused-code ratio — tree-shaking and route-splitting recover the wasted bytes.
Source: web.dev
Google Tag Manager takes 520ms of CPU time. Consider loading it asynchronously or replacing it with a lighter alternative.
This script has high main-thread execution time — optimize hot paths or defer.
Source: web.dev
cdn-cookieyes.com takes 508ms of CPU time. Consider loading it asynchronously or replacing it with a lighter alternative.
This script has high main-thread execution time — optimize hot paths or defer.
Source: web.dev
cohere.com takes 390ms of CPU time. Consider loading it asynchronously or replacing it with a lighter alternative.
This script has high main-thread execution time — optimize hot paths or defer.
Source: web.dev
59% of cohere.com's code is unused. The script may be loading features you don't use.
Bundle has high unused-code ratio — tree-shaking and route-splitting recover the wasted bytes.
Source: web.dev
t.contentsquare.net takes 383ms of CPU time. Consider loading it asynchronously or replacing it with a lighter alternative.
This script has high main-thread execution time — optimize hot paths or defer.
Source: web.dev
78% of t.contentsquare.net's code is unused. The script may be loading features you don't use.
Bundle has high unused-code ratio — tree-shaking and route-splitting recover the wasted bytes.
Source: web.dev
Google Tag Manager takes 359ms of CPU time. Consider loading it asynchronously or replacing it with a lighter alternative.
This script has high main-thread execution time — optimize hot paths or defer.
Source: web.dev
cohere.com takes 311ms of CPU time. Consider loading it asynchronously or replacing it with a lighter alternative.
This script has high main-thread execution time — optimize hot paths or defer.
Source: web.dev
cohere.com takes 307ms of CPU time. Consider loading it asynchronously or replacing it with a lighter alternative.
This script has high main-thread execution time — optimize hot paths or defer.
Source: web.dev
59% of cohere.com's code is unused. The script may be loading features you don't use.
Bundle has high unused-code ratio — tree-shaking and route-splitting recover the wasted bytes.
Source: web.dev
53% of cohere.com's code is unused. The script may be loading features you don't use.
Bundle has high unused-code ratio — tree-shaking and route-splitting recover the wasted bytes.
Source: web.dev
94% of cohere.com's code is unused. The script may be loading features you don't use.
Bundle has high unused-code ratio — tree-shaking and route-splitting recover the wasted bytes.
Source: web.dev
90% of cohere.com's code is unused. The script may be loading features you don't use.
Bundle has high unused-code ratio — tree-shaking and route-splitting recover the wasted bytes.
Source: web.dev
83% of cohere.com's code is unused. The script may be loading features you don't use.
Bundle has high unused-code ratio — tree-shaking and route-splitting recover the wasted bytes.
Source: web.dev
52% of cohere.com's code is unused. The script may be loading features you don't use.
Bundle has high unused-code ratio — tree-shaking and route-splitting recover the wasted bytes.
Source: web.dev