Open source at the core. Simple prices on top.
Adopt the core free, like Playwright. Pay us when you want your flows kept green on our infrastructure instead of yours.
Pre-launch: these plans go live with billing soon. The open-source core is free now; hosted tiers open in waves with $20 of credit to start, no card required.
Open Source
The full product, self-hosted on your own hardware.
$0 forever
Bring your own compute and model, or run our free local model backend.
- Everything: recorder, self-heal, deterministic Playwright codegen
- The CLI and website-to-API
- All run modes: test, data-entry, scrape
- The agent, on your own model
- Unlimited seats and flows, your infra
Open-source core, self-hostable. Public repo opens with launch.
Solo
Hosted, for one person keeping their own flows green.
$20 /mo
1,333 browser-minutes included ($20 of browser time), then $0.015 per browser-minute.
Start with $20 in free credit: 1,333 browser-minutes, no card required
- $20 of browser time included every month
- Overage $0.015 per browser-minute, runs keep going
- One tenth of QA Wolf's published $0.15 per runner-minute
Credit applies when hosted billing opens.
Team
Hosted on our infrastructure, maintained by the agent.
$99 /mo
10,000 browser-minutes included, then $0.015 per browser-minute.
Start with $20 in free credit: 1,333 browser-minutes, no card required
- ~25 flows kept green by the agent
- Self-heal and maintenance included
- Vercel deploy checks
- 3 seats
- Email support
$99/mo, and your first $20 is on us. Credit applies when hosted billing opens.
Scale
For teams running QA across a whole product surface.
$499 /mo
60,000 browser-minutes included every month, then $0.015 per browser-minute.
- ~150 flows kept green by the agent
- Priority run queue with burst capacity
- Unlimited seats
- SSO
- SLA
Enterprise
Your infrastructure or ours, with us on the hook.
Custom
Volume pricing, shaped to your deployment.
- Self-host support
- Dedicated runners
- Volume pricing
- Security review
One meter, and it's not AI. We meter browser-minutes, the compute a real test run uses. We never meter AI: our agent runs on our own models, so there are no AI credits, token packs, or per-decision charges on any plan.
What's in each plan
| Open Source | Team | Scale | Enterprise | |
|---|---|---|---|---|
| Flows maintained by the agent | Unlimitedself-run, your model | ~25 | ~150 | Custom |
| Browser-minutes / month | Unmeteredyour hardware | 10,000 includedthen $0.015/min | 60,000 includedburst capacity | Custom |
| Seats | Unlimited | 3 | Unlimited | Unlimited |
| Vercel deploy checks | Self-configured | Included | Included | Included |
| Self-heal + maintenance | Includedruns on your infra | Included | Included | Included |
| Run modes (test / data-entry / scrape / API) | All four | All four | All four | All four |
| Self-host | Yes, it's the product | Hosted by us | Hosted by us | Yes, supported |
| SSO | Your infra, your auth | Not included | Included | Included |
| Support | Community | Priority, with SLA | Dedicated, plus security review |
Most QA-as-a-service is enterprise-priced and per-flow. QA Reef is open-source at the core, and its hosted overage is $0.015 per browser-minute, one tenth of QA Wolf's published $0.15 per runner-minute: the division is shown, and the compare page cites what each vendor actually publishes, including where a rival beats us. Sizing this for a team with no QA function: QA without a QA team works the minutes out against a small suite.
Questions
What is a browser-minute?
One minute of a real browser running your flow on our infrastructure, rounded up per run. It is the compute a test run actually uses, so it is the only thing we meter. Recording, editing, and reviewing results cost nothing.
Is the core really free?
Yes, $0 forever. The open-source core is the full product: recorder, self-heal, deterministic Playwright codegen, the CLI, website-to-API, all four run modes, and the agent. You bring your own compute and model, or run our free local model backend. The paid plans sell our infrastructure and our upkeep, not features held back from the core.
Can I self-host?
Yes. Self-hosting the open-source core on your own hardware is the intended default, the same way teams adopt Playwright. On Enterprise we additionally support your self-hosted deployment directly, with dedicated runners available.
What happens when I hit the included browser-minutes?
Usage past the included browser-minutes (1,333 on Solo, 10,000 on Team, 60,000 on Scale) bills at $0.015 per browser-minute on every plan; your runs keep going and nothing is throttled. Scale adds burst capacity on a priority queue. You can see minute usage per run, so the number is never a surprise.
Do you meter AI usage?
No. We meter browser-minutes, the compute a real test run uses. We never meter AI: our agent runs on our own models, so there are no AI credits, token packs, or per-decision charges on any plan.
How is this cheaper than QA Wolf?
QA Wolf's published self-serve rate card is 15¢ per runner-minute plus 1¢ per AI credit (as published at qawolf.com/pricing, 2026-08-30); its managed Coverage as a Service has no published price. Per minute, our $0.015 is exactly one tenth of their published $0.15. At plan volumes the fraction is smaller still: Team's 10,000 included minutes would be $1,500 at QA Wolf's published rate against our $99 flat, about one fifteenth, and Scale's 60,000 would be $9,000 against our $499, about one eighteenth, before any AI credits. The full division, with assumptions, is on the QA Wolf pricing page.
Is there a hosted free tier?
Yes. The hosted free tier includes 500 browser-minutes a month, and new signups also get $20 in credit (1,333 browser-minutes at the overage rate), no card required. The self-hosted open-source core is free without any minute cap.