Plan philosophy

This is a companion to the plan tier limits reference. The reference tells you what each tier includes; this page explains why.

The principle

A pricing model that throttles you at the worst moment — your build’s red, you’re trying to debug, you hit the cap — is a hostile one. So the design rule is: Free should fence by capability, not starve you on volume, and each paid tier should earn its price with features you can feel, not just bigger numbers.

Free is fenced, not starved

Free gives you 300 cloud runs a month — enough to actually evaluate the product, not a 100-run trial that dies on a single CI job. What Free doesn’t give you is the serious-team capabilities:

Chromium only. Cross-browser (Firefox, WebKit) is the Starter line.
No BYOK, no local agent. Both are paid hooks.
1 seat, 2 schedules. Enough to prove scheduling works; a real cadence is Starter.

The idea: you can fully evaluate on Free, but the moment you’re doing real cross-browser CI for a team you’ve crossed a fence — and the fence is a felt, legible reason to upgrade, not an arbitrary volume wall.

BYOK is a paid hook

BYOK — connecting your own OpenAI / Anthropic / Groq / Gemini key — is a Starter-and-up feature. It moves AI inference onto your own provider account (you pay the provider directly, at provider rates) — what teams with their own provider agreement or compliance needs want. It’s reserved for paid tiers because it’s a real differentiator, not a Free giveaway.

Parallel executions scale with tier

Free / Starter get 1 parallel execution per org. Pro 2, Team 5, Enterprise unlimited.

Parallelism is the most expensive cap to raise — each slot is a Playwright instance burning CPU — and it’s where heavy users feel real productivity gain (5 tests at once in CI is 5× faster). Paying tiers get more slots; Free is one-at-a-time.

Retention is bounded by storage reality

Run history retention: Free 14 days, Starter 90 days, Pro 180 days, Team 2 years, Enterprise unlimited.

Screenshots are the real storage cost, so retention is the cap that maps most directly to what it costs to keep your data around. Free’s two weeks is enough to notice a trend; Starter’s 90 days covers a release cadence; Pro’s 180 days covers a typical audit window; Team’s two years covers compliance needs without an unbounded storage tail. Enterprise gets genuinely unlimited retention as part of a contract.

Seats and team features pay for themselves

Shared projects and org roles (Owner / Admin / Member / Viewer) work on every multi-seat plan — Starter and up. Free has a single seat, so there’s nobody to share with yet.

Team is where larger-group collaboration becomes practical:

10 seats — room for a real QA + dev group (Starter has 2, Pro has 5)
Google Workspace SSO — one click to add a new hire

Enterprise adds SAML SSO, audit logs, and an SLA — things procurement asks about, not things individual users need.

The promise

The reference page lists the numbers; this page promises the spirit:

You start with everything. New accounts begin on a 14-day Pro trial (no card), then move to Free — which stays useful forever: fenced by capability, not by a countdown.
Caps line up with growth, not with arbitrary paywalls.
No silent cliffs. You see usage meters before you hit a limit.