Eagle Eye, free for any single page
Test any webpage for accessibility problems in seconds.
Real Chromium, real rendering, axe-core rules, plain-language explanations. A11yBird tells you what it found — and what it could not test.
Scans run at a viewport width of 1440 pixels, against WCAG 2.2 level AA, including axe-core best-practice rules. Five scans per day, no signup.
- pages known
- 146
- templates
- 9
- last render
- 6.2s
Sample report
northwind.example.com/pricing22 Aug 2026, 14:07 UTCChromium 140, axe-core 4.13Scan your own pageA11yBird score
78/100
- 3 critical
- 7 serious
- 5 moderate
- 2 minor
Automated results only. Not a conformance claim.
Flight path
Spot it, explore it, understand it, watch it.
- Free
Eagle Eye
One page, tested in a real browser. Every finding explained, with the fix and what a human still has to check.
color-contrast×9image-alt×4 - From $49
Flight
Sitemaps, indexes, internal links. A11yBird discovers your pages, dedupes them and tests what matters.
82 of 146 pages56% - Included
Bird's-Eye View
Findings grouped by component and template, ranked by how many pages one fix would clear.
footernavcards - From $29 a month
Chirp
Scheduled Flights, then one short message when something regresses, resolves or moves the score.
Regression09:1416 new keyboard findings, all from SiteNav.
The scale of it
The scale of the problem
95.9%
of the top million homepages had automatically detectable WCAG failures.
56.1
detected errors per homepage, on average.
56M
individual errors across the sample.
WebAIM Million, February 2026
Measured on the rendered DOM, after scripts and styles. Automated testing finds a subset of problems — no detected failures is not the same as accessible, and no scanner can prove compliance.
Remediation leverage
A scanner hands you rows. A11yBird hands you the cause.
2,184
raw violations
31
underlying problems
Seven appear site-wide. Fixing three shared components could resolve roughly 71% of everything A11yBird detected.
Footer link colour
Navigation
Product cards
Unique one-offs
Highest-impact fix
One shared CSS rule on .footer__link accounts for 864 findings across 487 pages.
70%
leverage
Chirp
Hear about it when something changes. Not every time nothing does.
Chirp runs your Flights on a schedule and sends one short message when a regression appears, a batch resolves, or we lose sight of your site. Cosmetic score movement and single-page flaps never make the cut — an alert you learn to ignore is worse than none.
- RegressionToday, 09:14
16 new keyboard findings, all from SiteNav. Introduced since Tuesday's Flight.
- CoverageMon, 03:20
112 pages started answering with a security check. We stopped scoring them, and we're telling you rather than reporting a clean site.
- ResolvedLast week
41 contrast findings cleared across 212 pages. One footer rule, as predicted.
Plans
Pricing
Scanning one page is free, always. Subscriptions are for the watching.
Scan any page free.No account, no card, 5 a day. Everything below is for watching a whole site over time.
Scan a page| Plan | Watch$29/mo | ProPopular$79/mo | Agency$349/mo | Agency Scale$699/mo |
|---|---|---|---|---|
| Sites, pages each | 1 site · 100 | 3 sites · 500 | 10 sites · 2,000 | 30 sites · 2,000 |
| Flight schedule | monthly | weekly | weekly | weekly |
| Chirps | coverage + critical | + regressions | + regressions | + regressions |
| Sub-accounts, white-label | — | — | included | included |
| Join the waitlist | Join the waitlist | Join the waitlist | Join the waitlist |
Pay annually and we include a full extra Flight — 100 to 2,000 pages depending on your plan. Not a discount; an extra audit.
Scanning a page is live and free right now. Flight, Bird’s-Eye View and Chirp are built and tested, but not yet open for signup — the prices above are what they will cost when they are.
Developers
Catch regressions before production.
Run A11yBird locally, in CI, or against staging. It reports what your diff introduced — not the debt you already know about.
- $ npx a11ybird https://localhost:3000
- - uses: a11ybird/scan@v1
- POST /v1/scans
2 new serious problems introduced by this diff. Blocking merge is optional.