SpeedCurve vs WebPageTest vs Calibre vs Lighthouse CI: Web Perf Monitoring Compared

Choosing a web performance monitoring tool is a meaningful decision. The right tool catches regressions before users notice them, gives engineers actionable debugging data, and fits naturally into team workflows. The wrong one gets ignored because it's too noisy, too manual, or too expensive to justify for the coverage it provides.

This guide compares four leading web performance tools: SpeedCurve, WebPageTest, Calibre, and Lighthouse CI. Each has distinct strengths — understanding those tradeoffs is how you pick the right one for your team.

Quick Comparison Overview

SpeedCurve

SpeedCurve was founded by Steve Souders (formerly of Google, Yahoo) and Mark Zeman, and reflects the deep web performance expertise its founders bring. It's built around the idea that performance monitoring should be continuous, visual, and tied directly to business outcomes.

What SpeedCurve Does Well

Continuous Synthetic monitoring: SpeedCurve runs automated browser tests on a schedule (hourly, every 6 hours, daily) from multiple global locations. This is the foundation of its value — you don't have to remember to run tests, and the trend data accumulates automatically.

Real User Monitoring: The LUX snippet collects field data from actual users, capturing Core Web Vitals as browsers report them. The combination of Synthetic + RUM in a single tool is rare and valuable — most competitors offer one or the other.

Competitive benchmarking: The ability to test competitor pages under identical conditions is a standout differentiator. Side-by-side Filmstrip comparisons and direct metric comparisons against competitors are features no other tool in this list offers.

Visual debugging: SpeedCurve's Filmstrip and waterfall views are polished and informative. The Filmstrip in particular is excellent for stakeholder communication — seeing exactly when your page becomes visually complete, frame by frame, is more convincing than any table of numbers.

Alerting and budgets: Performance budgets with Slack, email, and webhook notifications are fully integrated. The alerting is stable and well-designed for team workflows.

SpeedCurve Limitations

Price: SpeedCurve is a commercial product aimed at professional teams. It's not cheap for small teams or individual projects. Pricing is based on the number of URLs tested, test frequency, and number of profiles.

No self-hosting: SpeedCurve is a SaaS-only product. If your compliance requirements or network policies require running tests in your own infrastructure, SpeedCurve isn't an option.

RUM requires a JavaScript snippet: LUX is lightweight but requires a <script> tag on every page. For sites with strict content security policies or third-party script restrictions, this may require additional configuration.

When to Choose SpeedCurve

• You need continuous monitoring, not just PR-level testing
• You want Synthetic + RUM in a single platform
• Competitive benchmarking is strategically important
• Your team wants polished dashboards and alerting without building custom infrastructure
• You're monitoring 10+ URLs across multiple profiles

WebPageTest

WebPageTest is the web performance industry's foundational tool. Originally created at AOL and now maintained by Catchpoint, it's open source and available both as a free public service at webpagetest.org and as a self-hosted deployment.

What WebPageTest Does Well

Depth of diagnostic data: No other tool produces more granular test data. WebPageTest provides waterfall charts, connection view, request details, response headers, content breakdown, video comparison, and a Chrome DevTools trace — all for a single test run.

Flexibility: Test from dozens of global locations, on real devices (mobile, desktop), with customizable connection profiles, and with advanced scripting to test multi-step flows. You can inject custom JavaScript, set cookies, handle authentication flows, and test behind login walls.

Open source and self-hostable: WebPageTest's server and agent code are open source on GitHub. Organizations with strict data residency requirements or who need to test internal staging environments can run their own instance.

Free public service: The public instance at webpagetest.org is free (with rate limits) and requires no account for basic use. It's the lowest barrier-to-entry option for one-off performance testing.

API access: The WebPageTest API lets you trigger tests programmatically and retrieve structured results, enabling CI/CD integration.

WebPageTest Limitations

No continuous monitoring: WebPageTest is fundamentally a single-shot testing tool. You run a test, get results, and that's it. There's no scheduler, no trend tracking, no alerting. To use it for continuous monitoring, you have to build that infrastructure yourself.

No built-in alerting: WebPageTest doesn't notify you when performance degrades. You'd need external tooling to trigger tests on a schedule, store results, compare against baselines, and send alerts.

Steeper learning curve: The depth of WebPageTest's data is also its complexity. The waterfall chart has dozens of data points; the settings panel has scores of options. New users often find it overwhelming. SpeedCurve and Calibre present similar data in more approachable ways.

No RUM: WebPageTest only measures Synthetic performance. You have no visibility into what real users experience.

When to Choose WebPageTest

• You need one-off diagnostic testing for specific performance problems
• You're debugging a complex waterfall and need maximum diagnostic detail
• You need to self-host for network/compliance reasons
• You're testing authenticated flows or multi-step user journeys with scripts
• Budget is zero and you're willing to build your own monitoring infrastructure

Calibre

Calibre is a commercial web performance monitoring platform with a focus on clean design and team workflows. It competes directly with SpeedCurve, offering continuous Synthetic monitoring, performance budgets, and CI/CD integration.

What Calibre Does Well

Clean UX: Calibre is well-designed and approachable. The dashboard is clean, onboarding is fast, and the learning curve is gentler than SpeedCurve's more feature-rich interface.

CLI tool: Calibre provides a first-class CLI (calibre test) that integrates cleanly into CI/CD pipelines without writing custom API scripts. A single command runs a test and returns pass/fail against your configured budgets.

Performance budgets in CI: Calibre's CLI makes it easy to use performance budgets as hard CI gates. The integration with GitHub Actions, CircleCI, and other platforms is well-documented with ready-made examples.

Snapshot testing: Calibre's snapshot comparison lets you compare performance across two specific points in time — e.g., before and after a refactoring — with side-by-side metric comparisons.

Calibre Limitations

No RUM: Like WebPageTest, Calibre only offers Synthetic monitoring. There's no field data to tell you what real users experience.

No competitive benchmarking: You can only test URLs from your own account. There's no built-in way to monitor competitor performance under the same conditions.

Price: Calibre's pricing is competitive with SpeedCurve but not cheap. Per-URL pricing with test frequency limits means costs scale with coverage.

Smaller ecosystem: SpeedCurve has a longer history and more community resources, blog posts, and integration examples. Calibre's ecosystem is smaller.

When to Choose Calibre

• You want continuous Synthetic monitoring with clean UX
• CI/CD integration via CLI is a priority
• You don't need RUM or competitive benchmarking
• Your team responds better to simpler interfaces than SpeedCurve's feature depth

Lighthouse CI

Lighthouse CI is Google's open-source tool for running Lighthouse audits in CI/CD pipelines. It's part of the broader Lighthouse ecosystem and focuses exclusively on the CI use case.

What Lighthouse CI Does Well

Free and open source: Lighthouse CI has no licensing cost. If you're already running your own CI infrastructure, adding LHCI is essentially free beyond the compute cost of running tests.

Native Lighthouse integration: Because Lighthouse CI runs Lighthouse directly, it has access to the full Lighthouse report — Performance, Accessibility, SEO, Best Practices scores, and all the diagnostic opportunities Lighthouse identifies. Other tools that integrate Lighthouse do so via API calls; LHCI runs it directly.

Assertion-based CI gates: Lighthouse CI's assertion syntax lets you define precise gates:

# lighthouserc.yml
ci:
  assert:
    assertions:
      'categories:performance':
        - warn
        - minScore: 0.9
      'audits:largest-contentful-paint':
        - error
        - maxNumericValue: 2500
      'audits:cumulative-layout-shift':
        - error
        - maxNumericValue: 0.1

GitHub integration: The LHCI GitHub Action posts results as a PR comment with a score summary, and supports blocking PR merges via required status checks.

Self-hosted server: The LHCI server (optional) stores historical results and shows trends over time, turning the CI tool into a lightweight continuous monitoring system.

Lighthouse CI Limitations

Lab data only: Lighthouse CI produces synthetic measurements in your CI environment, not from real browsers in diverse locations. Results depend heavily on the CI machine's resources — a busy runner will produce worse scores than a dedicated test machine.

No scheduled monitoring: LHCI runs only when triggered by your CI pipeline. Between deploys, you have no visibility. If a third-party script degrades performance overnight, you won't know until the next PR.

No alerting: LHCI has no notification system. You see results in the CI log and PR comment, but there are no Slack alerts or email notifications when performance degrades.

No RUM, no competitive benchmarking: LHCI is purely a Synthetic, in-CI testing tool. It doesn't help with understanding real user experience or competitive positioning.

Result variability: Running Lighthouse in CI produces more variable results than a dedicated test environment. CPU throttling is simulated, not real, and CI machine load affects results. This makes strict metric-based assertions prone to false failures.

When to Choose Lighthouse CI

• You need zero-cost performance testing in CI
• You want full Lighthouse audit coverage (accessibility, SEO, best practices) not just performance
• Your team is already familiar with Lighthouse and its output format
• You don't need continuous monitoring between deploys
• You're comfortable with open-source tooling and self-management

Head-to-Head: Key Comparisons

SpeedCurve vs. Calibre

These are the closest competitors. Both offer continuous Synthetic monitoring, performance budgets, and CI integration. The key differentiators:

Choose SpeedCurve if: You need RUM in addition to Synthetic, competitive benchmarking matters, or you want the most feature-complete performance monitoring platform.

Choose Calibre if: You prefer a cleaner, simpler interface, the CLI-first CI integration is important, or you don't need RUM or competitive data.

SpeedCurve vs. WebPageTest

These tools serve different primary use cases.

SpeedCurve: Continuous monitoring, alerting, team workflows. The answer to "are we getting slower over time?"

WebPageTest: Deep, one-off diagnostic testing. The answer to "why is this specific page slow right now?"

The best teams use both: SpeedCurve for ongoing monitoring and alerts, WebPageTest for deep-dive debugging when monitoring surfaces a problem.

SpeedCurve vs. Lighthouse CI

SpeedCurve: Commercial, comprehensive, includes RUM and continuous monitoring, designed for team workflows with proper alerting.

Lighthouse CI: Free, open-source, CI-only, no continuous monitoring, no RUM.

These tools aren't really competing for the same use case. Lighthouse CI is a CI gate; SpeedCurve is a full monitoring platform. Large teams often use both: LHCI for PR-level gating (fast, free, covers Lighthouse's broader audit scope) and SpeedCurve for continuous monitoring and RUM.

When to Stack Tools

The tools in this comparison aren't mutually exclusive. Common stacks:

The enterprise stack: SpeedCurve (continuous monitoring + RUM) + WebPageTest (diagnostic deep dives)

The budget-conscious stack: Lighthouse CI (PR gating) + WebPageTest (occasional deep testing)

The complete CI stack: SpeedCurve (monitoring + alerts) + Lighthouse CI (PR gating for Lighthouse-specific audits)

Feature Deep Dive: Real User Monitoring

Only SpeedCurve offers RUM in this comparison. This is a significant gap for competitors when it comes to understanding actual user experience.

RUM matters because Synthetic tests — no matter how sophisticated — cannot capture:

• The distribution of real device performance (old Android phones, low-RAM laptops)
• Geographic variability in CDN performance
• The effect of user behavior (pre-fetched pages, browser cache state, extension interference)
• INP (interaction responsiveness under real usage conditions)

If your site serves a diverse global audience on a wide range of devices, RUM data will regularly surprise you — showing that pages which look fine in Synthetic are significantly slower for a large fraction of your users.

Feature Deep Dive: Competitive Benchmarking

SpeedCurve is alone among these tools in offering built-in competitive benchmarking. This feature deserves extra attention because its business value is often undersold.

Under identical test conditions (same browser, network profile, and test infrastructure), you can see exactly how your LCP, CLS, and load time compare to competitors. This data:

• Justifies performance investment ("our LCP is 2x worse than [Competitor]")
• Sets realistic targets ("they're achieving 1.5s LCP, so that's achievable")
• Provides early warning when competitors improve ("they dropped their LCP by 800ms last month — they shipped something significant")

None of the other tools in this comparison offer this natively. You could approximate it by running WebPageTest against competitor URLs and manually comparing, but you'd lose the continuity and automation.

Pricing Guidance

Exact pricing changes frequently, so check each vendor's website for current rates. General guidance:

• Lighthouse CI: Free (open source, self-hosted)
• WebPageTest: Free public service; paid API for higher volume; self-hosted for free
• Calibre: Per-URL pricing, monthly plans starting in the hundreds of dollars for small teams
• SpeedCurve: Per-URL pricing, monthly plans, scales with URL count and test frequency; tends to be in a similar range to Calibre

For solo developers and very small teams, Lighthouse CI + WebPageTest's free tier provides meaningful coverage at no cost. For professional engineering teams where performance directly impacts business outcomes, SpeedCurve or Calibre's SaaS pricing is typically justifiable given the time saved on manual testing and the value of catching regressions early.

Making Your Decision

Start with SpeedCurve if: You're a professional team that needs comprehensive, continuous performance monitoring with RUM, competitive intelligence, and reliable alerting — and you're willing to pay for a polished SaaS solution.

Start with Calibre if: You want continuous monitoring with a cleaner learning curve, don't need RUM, and value strong CLI/CI tooling.

Start with WebPageTest if: You need deep diagnostic testing on demand, have budget constraints, or need self-hosting. Pair with a monitoring solution for continuous coverage.

Start with Lighthouse CI if: You want to add performance gating to CI for free, you value Lighthouse's full audit scope (not just performance), and you don't need continuous monitoring between deploys.

Use multiple tools if: Your needs span monitoring, CI gating, and deep diagnostics — because no single tool excels equally at all three.

The most important thing is to start somewhere. Any of these tools, configured and actively used, is dramatically more effective than no performance monitoring at all.