Visual testing you can predict, on a price you can read
TestKase Visual compares every capture with the image your team approved, pixel for pixel, after masking the parts you told it to ignore. Same inputs, same answer — on one flat monthly plan for the whole organisation, with 1,000 free snapshots to start.
1,000 free snapshots · lifetime allowance · no credit card required
Why teams choose TestKase Visual
A comparison you can explain to anyone on the team
- A pixel comparison: same inputs, same answer, every time
- Ignore regions masked out of both images, so noise never reaches you
- A six-part baseline key, so a resize or a second browser is never a false red
- Capture from a page list, a Live session, a folder of PNGs or a schedule
- No SDK to add and no framework to be tied to
- 1,000 free snapshots, then one flat monthly plan — never per seat
Judge it on your own pages
1,000 snapshots free for life and no card. Add the pages you care about, approve the first capture, and see what the next one finds.
Every flag has a reason you can point at
Masks first, then pixels, then a person. Because nothing in the middle is a judgement, a flagged capture always means pixels outside your masks changed — and a quiet run means they did not.
- Ignore regions masked out of both images before comparing
- A plain difference ratio, not a confidence score
- Unreviewed until someone marks it Approved or Bug
Six values decide what a capture is compared with
Project, page, browser, operating system, width and height. Chrome on Windows never overwrites Chrome on macOS, and a capture at a new size starts its own baseline instead of lighting the board up red.
- Chrome, Edge, Firefox and Safari each keep their own image
- A resize is a new baseline, never a regression
- Approve once, and every later capture is measured against it
- Values in the baseline key: project, page, browser, OS, width, height
- 6
- Ways to get a capture in, only one of which needs a test suite
- 4
- Review states: Unreviewed, Approved, Bug
- 3
- SDKs or framework plugins to install
- 0
Our figures, from the Visual product pages on this site.
TestKase Visual and Applitools, side by side
Where TestKase Visual stands out for a team that wants screenshots watched without a new programme to run.
| Item | Side by side | Status |
|---|---|---|
| How a difference is decided | TestKase Visual compares the pixels outside your ignore regions and reports a difference ratio, so the result is reproducible and easy to explain. Applitools uses its Visual AI with four match levels. | Predictable |
| Getting a first result | A TestKase run takes a list of public page addresses and opens them on our machines — nothing to add to your repository. Applitools connects through native integrations for Playwright, Cypress, Selenium, WebdriverIO, Appium and Storybook. | No SDK |
| Watching a page nobody deployed | A TestKase page can carry its own daily, weekly or monthly schedule, so a content edit or a third-party script change is caught without anyone starting a build. | Built in |
| Seeing the price | Every TestKase price is on the pricing page, including the paid Visual plan. Applitools publishes its entry plan and quotes Professional and Enterprise through sales. | Published |
| Seats | Neither charges per user: every Applitools tier advertises unlimited users, and TestKase Visual is one plan for the whole organisation, with nothing per parallel session either. | Both |
| What it sits beside | Visual is one of five products in one TestKase workspace, next to the device farm, the automation runs, the accessibility scanner and the test cases. | One workspace |
One flat plan for the whole organisation
Start with 1,000 snapshots free for life and no card. When you need more, one flat monthly plan covers the whole organisation — so the designer, the product manager and the support lead can all open the review queue at no extra cost.
- A snapshot is one page, in one browser, at one size
- Nothing per user and nothing per parallel session
- Every price published, no sales call to find it
Four ways in, and only one of them needs a test suite
Start from whatever you already have — a list of pages, a person, a folder of screenshots, or a clock.
| Item | What it takes to start |
|---|---|
| A run over public pages | Give it a list of addresses and press Run. A real browser opens each one on our machines and photographs it. Runs do not sign in, so this covers marketing pages, documentation and anything else a logged-out visitor can reach. |
| A capture from a Live session | Open a Live session, sign in, get the page into the state you care about — three steps into a wizard, a modal open, an error showing — and press Visual in the session rail. This works from a phone as well as from a desktop browser. |
| A folder of PNGs from the CLI | The command-line uploader takes screenshots your own pipeline already produced, whatever produced them. It exits 0 when nothing needs review and 1 when something does, which is the entire CI integration. |
| On a clock, with nobody driving | A page can carry its own schedule — daily, weekly or monthly at a time in a named timezone, on desktop browsers, with the report emailed to whoever set it up. |
Visual testing next to the rest of your testing
The browser that captures your baseline is the same fleet that runs your suites.
TestKase Visual lives in the same workspace as the products it is actually related to. The real browsers that capture a baseline are the fleet that runs your Selenium, Appium, Playwright, Espresso and XCUITest suites. The Live session where you press Visual is the session where you file the bug. The accessibility scanner runs in a real desktop browser on those same machines.
The test case that says the pricing table renders correctly lives in the same project as the baseline that proves it — one login, one bill, and one place to look when something on the page moves.
Frequently asked questions
It compares the new capture with the approved image, pixel for pixel, after masking any ignore regions out of both, and reports a difference ratio. There is no model weighing what matters, so the same two images always give the same answer, and a flag always points at pixels that actually changed.
You draw a box over the part of the page that changes on its own — a rotating banner, a date, a live counter — and that box is masked out of both images before they are compared. What is covered is genuinely not checked, and everything outside it is checked at full strictness. Regions apply from the next capture onwards, so an image you already approved is never re-interpreted by a mask added later.
A baseline is keyed by six values: project, page address, browser, operating system, and the width and height of the captured image. Chrome on Windows and Chrome on macOS hold separate approved images, because the same page does not render identically on both. A capture at a different size is a different baseline rather than a failure, which is why a resized window never reads as a regression.
No. A run takes a list of public page addresses and opens them in real browsers on our machines. You can also capture any state from a Live session, or upload a folder of PNGs your own pipeline already produces with the command-line uploader. None of the three adds an SDK to your repository.
Yes. Open a Live session, sign in, get the page into the state you care about and press Visual — on a desktop browser or a real phone. Or have your own automated test produce the screenshots and upload them with the CLI. Scheduled runs open public addresses, so they suit marketing pages, documentation and anything a logged-out visitor can reach.
Every account gets 1,000 snapshots free as a lifetime allowance, with no card. The paid plan is one flat monthly price with 5,000 snapshots a month for the whole organisation. Nothing is charged per user or per parallel session, so adding a designer to the review queue costs nothing. Every price is on the pricing page.
Sign up, add the public pages you want watched, and run them. The first capture of each page becomes its baseline once you approve it, and every capture after that is compared with it. Add a daily, weekly or monthly schedule when you want it to run with nobody driving.
Related
- TestKase vs PercyBaselines, ignore regions and review, in the same workspace
- TestKase vs ChromaticWhole-page visual testing for any site, with no Storybook required
- TestKase vs Deque axeThe same engine underneath, a different report on top
- All of Why TestKaseHow we compare with the tools you are probably already paying for
- PricingPer-seat test management and per-thread device testing, priced on the page
See exactly what changed on your own pages
1,000 free snapshots, no credit card, and a comparison you can explain in one sentence.