MotionSpec vs MotionScore, MotionLint, Lighthouse, axe, stylelint-a11y

What each tool measures about UI motion — reduced-motion guards, WCAG 2.2.2 pause paths, flashing, runtime JavaScript — with prices and reach as we found them on 3 September 2026. Written by the MotionSpec team; read it as one vendor’s honest inventory, not a neutral review.

The short version. Nobody covers all of motion accessibility, and the tools overlap less than their names suggest. Linters catch a missing prefers-reduced-motion block in CSS. Runtime tools measure what actually moves in a browser. General accessibility engines such as axe and Lighthouse barely look at motion at all. MotionSpec sits in a narrow spot: a static scan of somebody else’s live site that returns a score, a badge and a dated report, plus a compiler whose output carries the reduced-motion fallback and a pause path by construction. Where a tool below does something MotionSpec does not, the table says so.

Side by side

Where MotionSpec is alone — and where it is not

• Alone: a static scan of a third-party live URL that returns a score, a badge and a dated, WCAG-cited report; a deterministic compiler whose output carries the prefers-reduced-motion fallback; explicit reporting of WCAG 2.2.2 pause paths for looping motion.
• Not alone, or not there: detecting a missing reduced-motion guard (stylelint-a11y, Impeccable, MotionScore and MotionLint all do it, in CSS or at runtime); flashing analysis (Impeccable and MotionScore cover it, MotionSpec does not); runtime measurement (MotionLint measures real frames, MotionSpec only discloses that a runtime library is present — and misses one that arrives in an external bundle).
• Everyone: none of these tools makes a site WCAG-, EAA- or ADA-compliant, and none replaces a full accessibility audit. Each covers a slice; the slices are listed above.

Which one to reach for

• You lint your own CSS in CI — stylelint-a11y gives you the reduced-motion rule for free; add MotionSpec’s motion audit if you also want the 2.2.2 pause-path check and a score.
• You want to know what actually moves in a browser — MotionLint measures frames at runtime; MotionSpec’s static scan cannot.
• You direct a coding agent — Impeccable’s playbook and the Anthropic frontend-design plugin instruct the model; MotionSpec’s MCP tools validate and compile the result against a reviewed catalog.
• You ship other people’s sites and need paper — the free check for the pre-flight, the Motion Report for the client file, the Agency plan for ten domains at once.
• You already run Lighthouse or axe — keep them; neither evaluates animation, prefers-reduced-motion, autoplay or an unpausable loop, so a motion tool sits beside them rather than replacing them. What the Lighthouse score does and does not tell you →

A note on MotionSpec’s own numbers

MotionSpec currently reports two scores for the same findings: the website’s free check and the Motion Report use scoring v2 (severity weights per root cause, log-dampened by occurrences, per-category caps); npx motionspec audit (v1.2.7) and the MCP tools still report scoring v1 (a flat deduction per finding). Same findings, different weighting — the same URL can show two numbers today. A unified score is planned. The reach figures for MotionSpec above are our own: 213 npm downloads a week and 0 GitHub stars on 2026-09-03.

How this table was made

Each row comes from the tool’s own documentation, source or npm/GitHub page as read on 2026-09-03, plus a run of the tool where one was possible. Prices are the published list prices on that day; where we only had a secondary source (axe DevTools Pro), the table says so. We will correct any cell a maintainer shows to be wrong — write to hello@motionspec.dev and the change lands with a date.

MotionSpec checks and compiles UI animation for AI-generated web apps — it does not generate AI video. Product names belong to their owners; this page is not endorsed by any of them.