Chromatic vs Applitools

Applitools publishes its own comparison against Chromatic and it ranks first for this term, so you have probably read their framing: pixel diffing with manual thresholds versus Visual AI. That framing is accurate and incomplete. The bigger difference is scope — Chromatic tests the stories you wrote, Applitools tests whatever your suite drives, plus mobile, desktop and PDFs.

At a glance

Applitools' own argument, assessed

Their comparison page makes four claims. Taking them in turn, honestly:

1. 01"Visual AI, not pixel diffing." Fair, and the most substantive point. Perceptual comparison genuinely avoids the threshold trade-off, where tight enough to catch a real break is also tight enough to fail on antialiasing.
2. 02"Test maintenance is automated" — grouping similar diffs into one review. Fair, and it matters more than it sounds on a wide matrix. Chromatic reviews each story and browser combination separately.
3. 03"Coverage beyond the browser." Accurate. Chromatic is browser-only; Applitools covers native mobile, desktop and PDFs, which is a category difference rather than a feature gap.
4. 04"Deployment options." Accurate. Chromatic is hosted SaaS only; Applitools offers dedicated cloud and on-premises.

What their page omits is the price gap — $179 against $667 billed annually, a 3.7x difference at entry — and the checkpoint split that makes page-level testing on the Starter tier much tighter than it appears.

The checkpoint split matters most here

Applitools Starter includes 100,000 component checkpoints or 1,000 page checkpoints. For a team weighing Chromatic, that distinction is decisive, because it maps directly onto what you are trying to test.

So the honest framing: if you are comparing these two for component coverage, Applitools' 100,000 checkpoints against Chromatic's 35,000 snapshots is a real volume advantage, partly offset by TurboSnap and heavily offset by price. If you are comparing them for page coverage, Chromatic is not a candidate at all and the comparison you actually want is Applitools vs Percy.

What Chromatic does better

Two things, both underrated.

Baselines. Chromatic's model lives alongside git history and persists through branch and merge, with squash and rebase detected via git provider APIs. It is the deepest baseline engineering in this category. The documented catch is that approving on main does not update other branches — fixed by the account-level preferMergedBaselines flag.

TurboSnap. Dependency-graph-aware change targeting that can bypass a build entirely for zero billed snapshots. Applitools' grouping reduces review effort; TurboSnap reduces capture volume, which is what you are billed for. Model it against your real git habits though — editing any file in a story's dependency graph prevents the bypass, global Storybook file changes force a full rebuild, and rebase or squash breaks the computation.

The question that settles it

Not "which is better" but "where do my regressions actually happen?"

If they happen inside components — a button, a card, a form field — Chromatic is cheaper, better integrated and sufficient. If they happen where components compose into real pages with real data, Chromatic will keep passing while your customers keep finding bugs, and no amount of perceptual comparison fixes a page nobody tested.

Both share the same blind spot: coverage is authored by hand. Chromatic needs a story; Applitools needs an SDK call. A route nobody wrote either for is invisible to both, which is the factor that most reliably ends a visual testing programme — and the one we built VisualMax to remove, so weigh that framing accordingly. Several tools now attack it from different angles — a whole category of tools now generates its own coverage.