Home / Compare
Visual regression testing tools compared
Six tools that catch visual changes, next to Screenery. Each one is good at something. This page says what, and where Screenery does a different job.
At a glance
Competitor facts and prices were checked in October 2026 on each vendor's own pages. Each tool's page links its sources.
| Tool | Best at | Who renders | Free tier | Lowest listed paid plan |
|---|---|---|---|---|
| Percy | Cross-browser rendering, real mobile devices | Percy's cloud, from DOM snapshots | 5,000 screenshots a month | $199 a month, billed annually |
| Chromatic | Storybook component testing and sign-off | Chromatic's cloud | 5,000 snapshots a month, Chrome only | $179 a month |
| Argos | Open-source review for Playwright and Storybook | Your CI | 5,000 screenshots a month, for personal repositories | $100 a month |
| Applitools | Visual AI matching, PDFs, regulated teams | Applitools' grid or your browser | None. A 14-day trial | $667 a month, paid annually |
| BackstopJS | Config-driven scenarios, self-hosted report | Your machine or Docker | Open source | None. Seeking a new maintainer |
| Lost Pixel | Stories, pages and custom shots in one run | Your CI | Open source | None. Archived April 2026 |
| Screenery | Stable screenshot URLs that follow your default branch | Your tests. Screenery stores what they wrote | A Free plan. Limits not published yet | Prices not published yet |
Where Screenery fits
Most tools on this page answer one question: did this pull request change the UI? Screenery answers that one too, then keeps going. Every accepted screenshot gets a URL you can paste into a README, a docs page or a help centre:
https://cdn.screenery.dev/acme/dashboard/checkout/payment-form@latest.png
@latest moves when your default branch pushes. A pull request gets its own @pr-N channel. Every push is stored as a version whose URL never changes. Like a package registry, you get immutable versions and tags that move. See How @latest moves and Screenshot URLs.
What Screenery adds on top of that:
- A gate before publishing. Point the Action at a baseline and a visual regression fails the job before anything is pushed. See Visual testing.
- A pull-request comment showing what changed against @latest. See Pull-request comments.
- An MCP server, so an agent can compare @pr-N with @latest and see what was added, removed or changed. See Agents.
- No stored secret on GitHub Actions. The Action trades GitHub's OIDC token for a five-minute credential. Argos offers OIDC uploads too, so this is not unique.
And what it does not do, so you can rule it out:
- It does not render pages. It stores the files your tests wrote, so there is no browser grid.
- It has no Storybook integration. If you test stories, look at Chromatic or Argos first.
- It has no approve button that resets a baseline. A baseline is a folder you commit, or the version a push replaces.
How to choose
- You test Storybook components: Chromatic, made by the Storybook maintainers, or Argos.
- You need Safari, Firefox and real phones: Percy or Applitools.
- You want an open-source review app: Argos.
- You run Playwright and want the screenshots in your docs, current on every merge: Screenery. Start with Getting started.
- You already use Playwright's toHaveScreenshot: keep it as your gate. Screenery can publish the goldens it accepted. See Playwright.
- You are on BackstopJS or Lost Pixel: both need a new home. The BackstopJS and Lost Pixel pages are migration guides.
One page per tool
-
Screenery vs Percy
Percy re-renders your pages in four browsers. Screenery publishes the screenshots your tests took.
-
Screenery vs Chromatic
The Storybook platform, and when a Playwright team needs something else.
-
Screenery vs Argos
The closest match. Share links that expire, against URLs that follow your default branch.
-
Screenery vs Applitools Eyes
Visual AI and enterprise controls, against pixel diffs and published URLs.
-
Moving from BackstopJS
Turn backstop.json scenarios into one Playwright test, then publish.
-
Moving from Lost Pixel
Keep your shots running after the archive, then move them to Playwright.
Docs and help-centre screenshot tools
These tools keep screenshots in help articles current. They do not test for visual regressions.
-
LaunchBrightly alternative
LaunchBrightly replays recorded recipes against your app and syncs approved images into your help centre. Screenery publishes the screenshots your Playwright tests take in CI.
-
Promptless alternative
Promptless Capture signs in to your app with a test account and re-takes docs screenshots when the UI changes. Screenery publishes the ones your tests already take.
-
Ferndesk alternative
Ferndesk re-takes screenshots inside its own hosted help centre. Screenery's URLs work in any docs tool.
FAQ
What is the best visual regression testing tool for Playwright?
It depends on the job. Argos is an open-source review app for Playwright and Storybook. Percy and Applitools render your pages in browsers you do not run. Screenery fits when you run Playwright and want the screenshots in your docs, current on every merge, with an optional gate that fails CI on a visual regression.
Which visual testing tools have a free tier?
Checked October 2026: Percy gives 5,000 screenshots a month, Chromatic 5,000 snapshots a month in Chrome only, and Argos 5,000 screenshots a month for personal repositories. Applitools has no free plan, only a trial. BackstopJS and Lost Pixel are open source. Screenery has a Free plan whose limits are not published yet.
Which visual testing tool should I use for Storybook?
Chromatic, made by the Storybook maintainers, or Argos. Screenery has no Storybook integration.