tldr: There are no public prices on Applitools' pricing page. Three plans exist: Starter with a free trial and 50 Test Units, Public Cloud, and Dedicated Cloud, and the paid two are quote-only annual contracts. Every plan advertises unlimited users and unlimited test executions, because the meter is somewhere else. Applitools charges by the Test Unit, and the quote is built on how many of them you need.
Applitools is the best-known name in visual regression testing. Its Eyes product compares screenshots with an AI engine that ignores rendering noise, and its Autonomous product generates and runs visual tests on its own. Both are sold through the same three plans, checked on Applitools' pricing page on 31 July 2026: Starter, Public Cloud, and Dedicated Cloud. Starter offers a free trial with 50 Test Units. The other two say "Custom" where a number would go.
Quote-only pricing is common in this market. What makes Applitools different is that the quote hinges on a metric most buyers have never priced before.
Everything is metered in Test Units
Every plan includes unlimited users and unlimited test executions. Read that twice, because it sounds like the expensive parts are free. They are, and they are not the meter.
A Test Unit is consumed differently depending on the product:
- In Autonomous, each monthly active test counts against your Test Units
- In Eyes, each page counts against your Test Units
Units are interchangeable between the two products, and contracts run annually with the option to extend.
The design is clever and it changes how you forecast. "Unlimited executions" means running your suite fifty times a day costs the same as running it once. But every distinct test you keep active, and every distinct page you check, consumes a unit each month. Growth in coverage, not growth in usage, is what raises the bill.
Forecasting your Test Unit count
Before the sales call, count two things.
For Eyes, count the distinct pages or screens your visual suite checks. A 40-page application checked across three viewports is not necessarily 120 units, but the page count is the driver, so get it precise.
For Autonomous, count the tests you expect to be active in a given month, not the tests you have ever created. A suite of 300 tests where 200 run in a typical month is closer to 200 units than 300.
Then compare your count against Starter's 50 free units. Small teams fit inside the trial tier for a while. Teams with hundreds of active tests should expect the Public Cloud conversation, and regulated teams that need isolation are steered to Dedicated Cloud.
Hidden costs beyond the quote
The license is not the loaded cost. Visual testing produces a review queue: every intentional UI change generates diffs that a human approves or rejects, and on a fast-shipping product that queue becomes a standing job. Applitools' AI reduces false positives from rendering noise, but it cannot know whether a redesign was intentional. Budget reviewer hours alongside the contract.
Baseline management is the other one. Every branch, viewport, and locale multiplies baselines, and stale baselines are the visual-testing version of flaky tests. The tooling comparison covers how the alternatives handle the same problem.
How Applitools pricing compares
Open-source options like BackstopJS cost nothing and catch pixel diffs, at the price of noisy comparisons and self-managed baselines. Percy publishes screenshot-based pricing that is easier to predict and less sophisticated in comparison quality. Applitools sits at the top of the category on capability and, by most third-party accounts, on price.
The structural alternative is not buying a visual testing line item at all. Bug0 Managed is $2,500 a month flat, and visual checks ride inside the end-to-end suite that a dedicated engineer builds and verifies on the Passmark engine, so there is no separate per-test meter to forecast. Whether that fits depends on whether visual regression is your whole problem or one slice of a broader coverage gap.
Is Applitools worth the cost?
If visual correctness is a product requirement, Applitools is the strongest dedicated tool, and the Test Unit model is friendlier than per-screenshot pricing for teams that run suites constantly. The trial tier makes the evaluation cheap.
Go into the quote knowing your unit count, because that number is the negotiation. And treat the reviewer time as part of the price, because the tool surfaces diffs but people still adjudicate them.
FAQs
How much does Applitools cost?
Applitools does not publish prices. Starter is a free trial with 50 Test Units. Public Cloud and Dedicated Cloud are annual contracts quoted by sales, sized primarily by Test Unit count. All plans include unlimited users and unlimited test executions. Checked 31 July 2026.
What is an Applitools Test Unit?
The consumption metric behind every plan. In Autonomous, each monthly active test consumes a unit. In Eyes, each page consumes a unit. Units are interchangeable between the two products. Executions do not consume units, which is why running your suite more often does not raise the bill but adding coverage does.
Does Applitools have a free plan?
Starter functions as one: a free trial with 50 Test Units, unlimited users, and unlimited executions. It is enough to evaluate the comparison engine on a real slice of your application before a contract conversation.
Why does Applitools use Test Units instead of per-user pricing?
Because in visual testing, users and executions are not where the cost lives. The expensive resource is the AI comparison across your active surface area. Metering active tests and pages ties the price to coverage, which is fairer for large teams and easy to underestimate for growing suites.
How does Applitools pricing compare to Percy or open-source tools?
Percy publishes screenshot-based prices that are cheaper and simpler, with a less capable comparison engine. Open-source tools like BackstopJS are free plus the engineering time to run them. Applitools is the premium option, and the gap widens as your active test count grows.
