Playwright itself is free and open source under the Apache 2.0 license, with no paid edition. The cost of running Playwright in 2026 comes from three places that the framework’s price tag never shows: the CI compute that runs your suite, an optional managed browser grid, and the tool you use to store and share results. Each one bills differently. This guide lays the paid Playwright ecosystem side by side and names what you actually pay for.
Is Playwright free to use?
Playwright is free. It’s an open-source framework from Microsoft, released under Apache 2.0, with no license fee, no seat cap, and no paid tier of the framework itself. The “playwright cost” you eventually pay is for infrastructure around it, not for Playwright.
What’s included for free
Everything the framework does is in the free package:
- All three browser engines (Chromium, Firefox, WebKit), installed with
npx playwright install - Parallel execution across workers and test sharding across machines
- The HTML reporter, trace viewer, and built-in retries
- The full API, codegen, and CI integration examples
There is no feature gate. A 5,000-test suite runs on the same free Playwright as a 50-test one.
Does Playwright have a paid version?
No. Playwright has no paid version, no Pro edition, and no enterprise SKU of the framework. Microsoft’s separate hosted service, Microsoft Playwright Testing, was retired on March 8, 2026, and Playwright Workspaces in Azure App Testing replaced it. That is a distinct product you opt into, not an upgrade to Playwright. You can run Playwright forever without paying Microsoft anything.
Where Playwright actually costs money
The real bill comes from running and operating the suite, not from the framework. There are two recurring costs almost every team hits, and a third that most budgets forget. How those costs are metered matters more than any sticker price: a real Gaffer customer running 7 projects uploads more than 1,100 test reports a week and pays $15/month with $0 in overage, where a per-test-result meter would scale that same load straight onto the bill.
CI compute is the largest recurring cost
Browser tests are slow and parallel, so a large suite burns build minutes fast. On GitHub Actions, for example, each plan includes a monthly compute allowance set in dollars, usage beyond it is billed per minute by runner type, and Playwright sharding multiplies that by the number of parallel machines you run. Standard runners on public repos and self-hosted runners don’t draw on it. Otherwise this cost applies whether or not you ever buy a Playwright-adjacent service, and it scales with suite size and run frequency. Check your CI provider’s current per-minute rate; it is the line item that grows first.
Managed cloud testing grids
If your own CI runners can’t supply enough parallel browsers, a managed grid runs them for you. Playwright Workspaces in Azure App Testing is the current first-party option. It rents browser capacity, so the bill follows usage rather than a flat subscription.
Microsoft Playwright Testing pricing
Microsoft Playwright Testing is no longer sold. It was retired on March 8, 2026, the date Microsoft’s docs set. Azure notice AZ499860 (third-party copy), published September 9, 2025, says remaining resources stop working after that date, and Microsoft publishes a migration guide. Its replacement is Playwright Workspaces in Azure App Testing, which runs Playwright tests on cloud-hosted browsers.
The notice says Azure App Testing has identical pricing to the service it replaced, so the meter carried over: you pay per minute that tests run in cloud browsers, through Azure.
How per-test-minute billing works
Microsoft’s pricing page defines test minutes as the total time tests run in its cloud browsers, billed per second: 50 tests averaging 12 seconds each is 10 test minutes. Running them in parallel finishes sooner but leaves that total unchanged.
What’s included in the Playwright Workspaces free trial
The Playwright Workspaces free trial is 30 days and 100 browser minutes, then pay-as-you-go. Rates are on Microsoft’s pricing page.
Currents.dev pricing for Playwright
Currents is a results-and-orchestration service for Playwright and Cypress that bills per test-result. You pay for the number of individual test results you upload, so the meter tracks suite size multiplied by how often CI runs, not browser compute. It’s a different axis than Microsoft’s grid: Currents doesn’t run your browsers, it ingests and orchestrates the results your CI produces.
How per-test-result billing works
A test-result is one test’s outcome in one run. A 1,000-test suite produces 1,000 results per run, and retries of the same test count as a single result. Currents sells monthly tiers defined by an included test-result quota, with overage once you exceed it. Because the quotas and price points change, check the current Currents pricing page for the exact tier sizes; the model to understand is that your bill tracks tests x runs, so a small suite run hundreds of times a day and a large suite run a few times a day can land at very different costs.
Estimating the bill
To estimate a per-test-result bill, multiply your test count by your CI runs per month:
results/month = test_count x runs_per_monthA 1,200-test suite run 200 times a month is 240,000 results per month. Whether that’s cheap or expensive depends entirely on which quota tier it lands in, which is why the per-result model is hard to predict for teams whose run frequency varies.
Playwright reporting and test storage: what it costs
Storing and sharing Playwright results is the cost most budgets miss, because the free OSS tooling doesn’t cover it. Playwright’s HTML reporter writes a report to disk; it does nothing to host that report at a URL, keep history across runs, or let a teammate without CI access open it. Closing that gap is a separate tool with its own pricing model.
Why you need a separate tool to store and share results
The Playwright HTML report is a folder of static files. To make it a link a colleague can click, you have to host it somewhere, because CI artifacts are downloads behind a CI login. To answer “is this test flaky or did it actually break?” you need history across runs, which a single report can’t show. Teams either wire up an artifact bucket and a static host themselves, or use a hosting platform that ingests the report and keeps the history.
Flat-rate vs metered pricing for test-result storage
Result-hosting tools split into two pricing models. Metered tools (per test-result, per seat) get more expensive as your suite grows, your team grows, or your run frequency climbs. Flat-rate tools don’t count uploads, test results, or viewers, though a plan can still bill for stored bytes over its limit.
Gaffer is the flat-rate option. It hosts Playwright reports at stable URLs, builds flaky-test detection and pass-rate trends across runs, and shares results by link without a login. Pricing is a flat monthly fee with unlimited users on every plan, plus storage overage on Pro and Team:
- Free: $0/month, 500 MB storage, 7-day default retention
- Pro: $15/month, 10 GB storage, 30-day default retention
- Team: $49/month, 50 GB storage, 90-day default retention
Storage overage on Pro and Team is $0.50/GB/month. The bill doesn’t count uploads or test results, but more uploads mean more stored bytes, and bytes over the plan limit are billed. The 7-project customer above stays under the 10 GB limit, so the bill has no metered line item.
Playwright pricing compared: OSS vs cloud grid vs results platform
These products solve different problems and bill on different axes, so a like-for-like price comparison is the wrong question. The right question is which cost you’re trying to control: browser compute, result volume, or sharing and history. The table below maps each.
Side-by-side: Playwright Workspaces vs Currents vs TestDino vs Gaffer
| Service | What it does | Billing model | You pay per | Free tier |
|---|---|---|---|---|
| Playwright (OSS) | The test framework itself | Free, open source (Apache 2.0) | Nothing | Entirely free |
| Playwright Workspaces | Managed cloud browsers (runs your suite). Replaced Microsoft Playwright Testing, retired March 8, 2026 | Pay-as-you-go, via Azure | Usage, at Microsoft’s rates | Trial: 30 days and 100 browser minutes |
| Currents.dev | Result ingestion + CI orchestration | Per test-result, monthly tiers | Each uploaded test result (tests x runs), with retries of the same test counted once | No free tier; 14-day free trial with up to 10,000 results |
| TestDino | Playwright reporting + test-case management | Tiered quota, plus overage on paid plans | Tier (execution quota and seat limit), plus $3 per 1,000 extra executions on Pro and Team (30 users on Team) | Free: 5,000 executions/mo, 1 user |
| Gaffer | Test-result hosting, sharing, flaky/trend analytics | Flat monthly rate, plus storage overage on Pro and Team | The plan, plus stored bytes over its limit | Free plan: $0, 500 MB, 7-day retention, unlimited users |
Confirm current rates on Microsoft’s Playwright Workspaces pricing page and Currents’ pricing page before budgeting; their numbers move. Gaffer’s are listed in full on the pricing page.
What should you budget for Playwright in 2026?
Budget for the framework at $0, then add the costs that match how you run it. If your own CI runners handle the parallelism, you skip a managed grid entirely and your only Playwright-specific spend is result hosting. If your suite is large or runs constantly, a per-test-result service can grow unpredictably, which is the case where storage-metered hosting can cost less if your reports are small. If you’ve already found per-result pricing painful, the Currents alternative comparison walks through the result-metered versus storage-metered tradeoff in detail.
The trap to avoid: budgeting for the compute and forgetting the storage. A team can pay for CI minutes to run the suite and still have no link to share and no history across runs. That second cost depends on how much you store, and it is an annoying scramble if you don’t plan for it.
FAQ
Is Playwright free or paid?
Playwright is free. It’s open source under Apache 2.0 with no paid version of the framework. Costs come from CI compute, optional managed browser grids, and result hosting, not from Playwright itself.
Does Playwright have a paid version?
No. There is no paid edition, Pro tier, or enterprise SKU of the Playwright framework. Microsoft’s separate hosted service, Microsoft Playwright Testing, was retired on March 8, 2026, and Playwright Workspaces in Azure App Testing replaced it. That’s a distinct product, not an upgrade to Playwright.
How much does Microsoft Playwright Testing cost?
Microsoft Playwright Testing is no longer sold: Microsoft retired it on March 8, 2026, and Playwright Workspaces in Azure App Testing replaced it. Playwright Workspaces bills per minute that tests run in its cloud browsers, through Azure. It has a free trial of 30 days and 100 browser minutes, then moves to pay-as-you-go; rates are on Microsoft’s pricing page.
How much does Currents cost for Playwright?
Currents bills per test-result, sold in monthly tiers with an included quota and overage beyond it. Your bill tracks suite size multiplied by run frequency (tests x runs), not browser compute, and retries of the same test count as a single result. As of November 5, 2026, there is no free tier; a 14-day free trial covers up to 10,000 results, and Scale is $49/month for 10,000 results. Check the current Currents pricing page for the exact tier sizes.
What’s the cheapest way to store and share Playwright reports?
For a small or growing team with small reports, flat-rate hosting is often cheapest because the bill doesn’t count uploads, test results, or headcount; on paid plans, stored bytes over the plan limit are billed. Gaffer’s free plan hosts reports at $0 with unlimited users; the Pro plan is $15/month flat, plus $0.50/GB/month for storage over its 10 GB limit. Per-test-result and per-seat tools get more expensive as the suite and team grow.
Playwright costs nothing to license. The spend is in the infrastructure around it, and the paid options bill on different axes: usage for a cloud grid, per test-result for orchestration, tiered quotas plus overage for reporting and test-case management, flat-rate plus storage overage for hosting and history. Match the model to the cost you’re trying to control. Gaffer’s pricing is on the pricing page.