Percy vs Applitools vs Chromatic: Visual Regression Testing Compared
Every time you ship a new feature, you are also shipping a risk: a CSS change that moves a button three pixels to the left, a font-weight update that cascades across six components, a responsive breakpoint that breaks on a viewport you did not test. Visual regressions are the bugs that automated unit tests and functional tests miss entirely — because they test behavior, not appearance.
Visual regression testing is the practice of automatically comparing screenshots of your UI before and after a code change, flagging anything that looks different for a human to review. It is the closest thing to having a QA engineer stare at every page, on every browser, on every merge.
Three tools dominate this space: Percy, Applitools, and Chromatic. They all solve the same core problem, but they approach it differently — different diffing technology, different workflow assumptions, different pricing, and different trade-offs. This comparison will walk through each one in detail so you can make an informed choice for your team.
What Visual Regression Testing Actually Is
The concept is straightforward. Before a change ships, the tool captures a baseline screenshot. After the change, it captures a new screenshot. It then diffs the two — pixel by pixel, or with AI — and surfaces the differences for a reviewer to accept or reject.
The hard part is not the comparison. It is the noise. Browser rendering engines render the same CSS slightly differently between versions. Anti-aliasing shifts individual pixels. Fonts render at sub-pixel variations across operating systems. A naive pixel-by-pixel diff of a production UI will flag hundreds of false positives on every run — not because something broke, but because Chrome 130 renders a particular glyph one pixel differently than Chrome 129 did.
How each tool handles this noise is, in many ways, the most important thing to understand about it.
Percy: CI-First Visual Testing with a New AI Layer
Percy, now owned by BrowserStack since 2020, is built around a simple premise: visual testing should live inside your CI/CD pipeline, require minimal setup, and give your team a straightforward review workflow. It integrates with Playwright, Cypress, Selenium, TestCafe, WebdriverIO, and Storybook, covering most of the test framework landscape.
How Percy Works
Percy injects itself into your existing test suite. As tests run, Percy captures DOM snapshots — not raw screenshots. It sends those snapshots to its cloud infrastructure, which renders them across browsers and viewports and generates the actual screenshots. This rendering-on-upload approach means you are not dependent on the browser running in your CI environment to produce consistent screenshots.
Comparisons are made against a baseline tied to your Git branch. When you open a pull request, Percy compares your branch's snapshots against your main branch baseline and surfaces a visual diff inside your PR checks.
Pixel Diffing vs the Visual Review Agent
Historically, Percy used pixel-level comparison with some noise filtering — an OCR library to handle minor text rendering shifts and a configurable diff threshold to ignore trivial changes. It was accurate but still produced noisy output on complex UIs. A font rendering change or a subtle anti-aliasing difference could highlight an entire section of a page in red, even though nothing meaningful had changed.
In late 2025, Percy shipped its Visual Review Agent — the most significant capability addition since the BrowserStack acquisition. The Visual Review Agent replaces raw pixel highlighting with AI-driven smart highlights. Instead of marking every changed pixel, it draws bounding boxes around meaningful changes and ignores the rest. The practical output shifts from "47,000 pixels changed" to "the header navigation shifted 4px left and the hero image was replaced." Percy reports that the Visual Review Agent reduces review time by approximately 3x and automatically filters out roughly 40% of visual changes as non-meaningful noise.
Percy Pricing
Percy has one of the most generous free tiers in this space: 5,000 screenshots per month with unlimited team members. For teams with higher volume, Percy's paid plans are available through BrowserStack. Parallelization — running multiple snapshot jobs simultaneously to speed up test execution — costs extra on most plans, which is worth factoring in if you have a large test suite and care about CI speed.
Percy's Best Fit
Percy works best for teams that need broad CI integration across multiple test frameworks and full-page visual testing, not just component-level snapshots. Its free tier makes it accessible for small teams and open source projects. The Visual Review Agent makes it meaningfully more useful for large codebases where pixel diff noise was previously the bottleneck in review workflows.
Applitools: Enterprise Visual AI at Scale
Applitools Eyes is the most technically sophisticated of the three tools, built around a visual AI engine that has been trained on millions of UI screenshots. Where Percy and Chromatic started as screenshot comparison tools and added AI later (or not at all), Applitools built AI-driven comparison as its core differentiator from the beginning.
How Applitools Works
Applitools uses what it calls Visual AI — a perceptual comparison engine that understands UI semantics rather than raw pixels. It recognizes that a button is a button, that a navigation bar has a structural relationship to the rest of the page, and that a 2-pixel shift in a UI element may or may not be a meaningful regression depending on context. The engine supports multiple match levels — Strict, Layout, Content, and Exact — letting teams tune how aggressively differences are flagged depending on what they care about testing.
Applitools integrates with virtually every major test framework: Selenium, Playwright, Cypress, WebdriverIO, Appium for mobile, and more. It runs on Applitools' Ultrafast Test Cloud, which parallelizes snapshot capture to reduce test execution time significantly.
The Figma Plugin and Storybook Addon
In January 2026, Applitools shipped Eyes 10.22 with two notable additions that expand where visual testing can happen.
The Figma Plugin lets designers export frames — individual components, pages, or full prototypes — directly into Applitools Eyes as visual baselines. Developers then run their standard visual tests against those design baselines. Eyes' Visual AI flags meaningful differences between the built implementation and the approved design, closing the gap between design intent and production reality. The plugin also supports design-to-design comparisons, so teams can diff two Figma versions against each other to track visual evolution across iterations.
The Storybook Addon brings component-level visual testing directly into the Storybook environment, allowing teams to run Eyes tests on individual stories without a separate test file. This gives Applitools an answer for component-library workflows that previously made Chromatic the obvious choice.
Applitools Pricing
Applitools is the most expensive option in this comparison. Pricing starts at approximately $969 per month, with enterprise packages that scale significantly beyond that. Applitools does not publish its full pricing publicly — enterprise teams typically go through a sales conversation to get a quote. There is a free tier, but it is limited to 100 screenshots per month, which is more a trial than a workable plan. For most teams, Applitools is a serious budget commitment.
Applitools' Best Fit
Applitools is built for large engineering organizations that need maximum accuracy, scalability, and sophisticated baseline management across complex applications and multiple platforms. Its Visual AI is the most mature in this space. The Figma plugin is a genuine differentiator for design-system teams that want to close the design-to-code validation gap. For smaller teams or projects with modest visual testing needs, the cost and setup complexity are difficult to justify.
Chromatic: Visual Testing Built for Storybook
Chromatic is built by the team behind Storybook, and that lineage shapes everything about how it works. If Percy is CI-first and Applitools is AI-first, Chromatic is Storybook-first. It is designed for teams that build UI as isolated components, develop with Storybook as their environment, and want visual regression testing that slots directly into that workflow.
How Chromatic Works
Chromatic captures snapshots of your Storybook stories — one snapshot per story, per browser, per viewport combination. It compares new snapshots against baselines and flags changes in a review UI where team members can accept or reject them. Beyond Storybook, Chromatic also supports Playwright and Cypress for full-page and interaction testing.
Chromatic's diffing is pixel-based. It does not have an AI layer comparable to Percy's Visual Review Agent or Applitools' Visual AI. What it offers instead is intelligent baseline tracking: baselines are tied to Git history, mirroring how developers already track code changes. Baselines persist through branches and merges without the baseline conflicts that can occur in tools with simpler branch-comparison logic. For teams working with multiple feature branches in parallel, this baseline management approach is a genuine practical advantage.
Speed and Parallelization
Chromatic runs test snapshots with unlimited parallelization on all plans — every story is captured simultaneously. The result is fast turnaround: Chromatic claims it can snapshot and diff 2,000 tests in under two minutes. This is a meaningful differentiator against Percy, which charges extra for parallel test runs, and Applitools, which does not publish parallelization details.
Chromatic Pricing
Chromatic's Starter plan is $179 per month; Pro is $399 per month. Both are meaningfully cheaper than Applitools for most team sizes. A free tier is available. Chromatic is competitively priced for teams already using Storybook, but the value proposition drops if your team is not Storybook-first — the tool's strengths are deeply tied to that component-driven workflow.
Chromatic's Best Fit
Chromatic is the right choice for frontend teams that build component libraries, use Storybook as their primary development environment, and want visual regression testing that feels native to that workflow. It is not a good fit for teams without Storybook, for full-page visual testing of complex user flows across arbitrary application states, or for teams that need AI-powered noise reduction on high-volume snapshot runs.
Head-to-Head: Key Dimensions
Diffing Technology
Applitools has the most mature AI visual comparison engine, trained on a large corpus of UI screenshots and capable of semantic-level understanding. Percy's Visual Review Agent is newer but delivers meaningful AI-driven noise reduction and smart change summarization. Chromatic uses pixel-based diffing with no AI layer — accurate but susceptible to false positives from rendering noise on complex UIs.
CI/CD Integration
All three tools integrate with major CI systems: GitHub Actions, GitLab CI, Jenkins, CircleCI, and others. Percy has the broadest framework support across Playwright, Cypress, Selenium, Storybook, and more. Applitools supports the widest range of platforms including mobile. Chromatic's integrations are strong but most naturally suited to Storybook-centric workflows.
Storybook Support
Chromatic is built by the Storybook team and offers the deepest, most seamless integration. Applitools now offers a dedicated Storybook Addon as of Eyes 10.22. Percy supports Storybook as one of several integration options. If Storybook support is your primary criterion, Chromatic is still the most natural fit, but the gap has narrowed.
Baseline Management
Chromatic's Git-based baseline tracking is the most developer-friendly approach, handling branch merges and parallel development naturally. Percy and Applitools both compare feature branches against a main branch baseline, which can create friction in complex branching workflows. This is a practical concern more than a dealbreaker, but worth understanding before committing.
Pricing Summary
| Tool | Free Tier | Entry Paid Plan |
|---|---|---|
| Percy | 5,000 screenshots/month | BrowserStack enterprise pricing |
| Applitools | 100 screenshots/month | ~$969/month |
| Chromatic | Available | $179/month (Starter) |
Which Tool Should You Choose?
There is no universally correct answer. The right tool depends on your team's workflow, scale, and budget.
Choose Percy if you want broad CI integration across multiple test frameworks, full-page visual testing, and a generous free tier that works for real projects. The Visual Review Agent makes it substantially more useful for teams that were previously drowning in pixel diff noise.
Choose Applitools if you are a large engineering organization with a significant budget, complex multi-platform testing requirements, and a need for the most accurate AI-powered visual validation available. The Figma plugin is a genuine advantage for design-system teams that want to validate design-to-code fidelity. Applitools is named a Strong Performer in the Forrester Wave for Autonomous Testing Platforms, Q4 2025 — it has the enterprise credibility to match its enterprise price.
Choose Chromatic if your team is Storybook-first, you build and maintain a component library, and you want a visual testing tool that slots natively into that workflow. The Git-based baseline tracking and unlimited parallelization are real practical advantages. If you are not using Storybook, Chromatic is probably not the right starting point.
Some teams use more than one: Percy for full-page CI testing and Chromatic for daily component-level development workflows. The tools are not mutually exclusive if your testing coverage requirements justify the cost.
Where Automated Visual Regression Testing Has Limits
All three tools are automation-first. They are designed to catch regressions introduced by code changes — visual differences between the state before a commit and the state after. They are not designed for the exploratory, manual phase of QA: the work of actually using a product, finding unexpected edge cases, and reporting visual bugs that live outside the scope of your test suite's coverage.
Automated visual regression testing covers the paths you thought to test. Manual QA covers the paths you did not. A pixel shift in your header navigation will be caught by Percy in your CI pipeline. A visual bug in a modal that only appears after a specific sequence of user interactions, on a viewport you did not write a test for, may not be.
This is where a tool like Crosscheck fits. Crosscheck is a Chrome extension built for the manual QA workflow — the exploratory testing, the edge cases, the bugs that fall between the coverage gaps of your automated suite. When a QA engineer finds a visual bug during manual testing, Crosscheck lets them capture an annotated screenshot or screen recording, automatically attaches the console logs and network requests from that moment, and files a rich bug report to Jira or ClickUp in seconds.
Crosscheck's Instant Replay feature retroactively captures the last one to five minutes of a session as a DOM-based replay — so even bugs discovered unexpectedly during exploratory testing have a full reproduction path attached. Developers get everything they need to reproduce and fix the issue without a single follow-up message.
Visual regression testing with Percy, Applitools, or Chromatic and manual bug reporting with Crosscheck are not competing workflows. They are complementary layers of a complete visual quality strategy: one catches automated regressions at the code level, the other captures the visual bugs that only surface when a human is actually using the product.
The Bottom Line
Percy, Applitools, and Chromatic each solve visual regression testing well. The differences are in approach, scale, and cost. Percy offers the best balance of breadth and accessibility, with a newly meaningful AI layer. Applitools offers the most sophisticated visual AI for teams with the budget to match. Chromatic offers the most natural fit for Storybook-centric frontend teams.
The more important point is that automated visual regression testing — however well implemented — is one layer of visual quality assurance, not the whole picture. The teams that ship the fewest visual bugs are the ones that combine automated regression coverage with disciplined manual QA and the tooling to make manual bug reporting fast, complete, and actionable.
Want to close the gap between automated visual regression testing and manual QA? Try Crosscheck free — annotated screenshots, retroactive session replay, and auto-captured developer context in a single Chrome extension.
