Workspace-shared by default
Teammates and Projects draw from one visible pool. Budgets remain a Workspace concern, never a hidden per-user meter.
One shared credit pool for the QA work your whole Workspace needs.
A completed run costs credits only when its evidence is finalized. PASSED or FAILED, you get screenshots, recording, reproduction steps, Console and Network context. If Venkat cannot reliably tell, the result is ERROR—and the run costs zero credits.
PASSEDVerified journey and finalized proof. Credits apply.
FAILEDVerified product failure and finalized proof. Credits apply.
ERRORVenkat could not reliably tell. Zero credits.
No per-seat tax. No charge for projects sitting idle. Select a plan to change the live Paper Shader and estimate.
Preview pricing in USD. The cards describe plan entitlement, not current production availability. Use the readiness column below: Phase 3 Web Test remains DOGFOOD-PENDING; mobile, desktop, execution-path integrations and enterprise controls remain later-phase commitments.
M2M infrastructure, MCP-first control, 900+ integrations, native agentemail + agentphone, exportable memory and PR-flow scale—with honest readiness beside every promise.
| Capability | Hobby | Startup | Enterprise | Readiness |
|---|---|---|---|---|
| Capacity and collaboration | ||||
| Verified run creditsStandard evidence-complete web journeys | 200 | 1,000 | Custom | PRICING PREVIEW |
| Concurrent test runs | 1 | 10 | Custom | LATER PHASE |
| Projects | 2 | 5 | Custom | FOUNDATION LIVE |
| Unlimited membersNo seat tax; one Workspace billing boundary | ✓ | ✓ | ✓ | FOUNDATION LIVE |
| Platforms, authoring and integrations | ||||
| Machine-to-machine run infrastructureOne isolated agent session and immutable context per attempt | ✓ | ✓ | ✓ | FOUNDATION LIVE |
| AI test authoringSaved, ordered tests from product context | ✓ | ✓ | ✓ | PRODUCTION EVIDENCED |
| Web application testing | ✓ | ✓ | ✓ | P3 DOGFOOD |
| iOS + AndroidAppium-backed device journey proposal | — | ✓ | ✓ | LATER PHASE |
| Desktop / Electron | — | — | ✓ | LATER PHASE |
| PR-flow orchestrationGitHub, Slack and webhook triggers that scale with release volume | — | ✓ | ✓ | LATER PHASE |
| MCP-first control surfaceParity with the trusted REST and CLI contracts | — | ✓ | ✓ | AFTER P3 |
| 900+ integration catalogConnector reach without placing third-party credentials in the agent | — | ✓ | ✓ | LATER PHASE |
| Native agentemail + agentphoneEmail, SMS, OTP and phone journeys as first-class test capabilities | — | ✓ | ✓ | LATER PHASE |
| Exportable structured Markdown memoryPortable project knowledge; no vector database lock-in | ✓ | ✓ | ✓ | LATER PHASE |
| Enterprise and managed service | ||||
| SAML SSO, SCIM and audit log | — | — | ✓ | LATER PHASE |
| Custom evidence retention + encryption policy | — | — | ✓ | LATER PHASE |
| Priority support, SLA and onboarding | — | — | ✓ | CONTRACTED |
| Custom integrations + strategic partnership | — | — | ✓ | CONTRACTED |
| Managed coverage + failure investigation | — | — | Scoped | PROPOSAL |
| Venkat proof layer | ||||
| ERROR costs zero creditsMissing evidence or uncertain cleanup can never become PASSED | ✓ | ✓ | ✓ | P3 CONTRACT |
| Screenshots, recording and reproduction steps | ✓ | ✓ | ✓ | P3 DOGFOOD |
| Separate Console + Network evidence | ✓ | ✓ | ✓ | P3 DOGFOOD |
| Ordered transcript survives reconnect | ✓ | ✓ | ✓ | P3 DOGFOOD |
| Immutable Run context + one isolated attempt | ✓ | ✓ | ✓ | FOUNDATION LIVE |
| Workspace-isolated reads fail closed | ✓ | ✓ | ✓ | PRODUCTION EVIDENCED |
| Evidence finalization + clean session termination | ✓ | ✓ | ✓ | P3 DOGFOOD |
Venkat shows the credit estimate before each Run. A normal saved web journey is one credit. Longer or exploratory work declares its estimate first—never after the bill.
Assumes one credit per standard evidence-complete web run. Advanced journeys use the estimate shown before execution. Taxes are excluded.
Credits buy a trustworthy answer—not animation, tokens, or agent time.
Teammates and Projects draw from one visible pool. Budgets remain a Workspace concern, never a hidden per-user meter.
The estimate is resolved from the saved Test and immutable Run context. The agent cannot choose the price or tenant scope.
A FAILED Run with finalized evidence is billable because it found a reproducible product problem. A Venkat ERROR is not.
Missing evidence, failed finalization, or cleanup uncertainty produces ERROR. Those Runs consume zero customer credits.
For this preview, one credit covers one standard saved web journey that completes with finalized evidence. Venkat displays a larger estimate before unusually long or exploratory work starts.
FAILED means the target product broke and Venkat proved it with reproducible evidence. That answer is the service. ERROR means Venkat could not establish a trustworthy answer, so the Run uses zero credits.
No. The model uses a Workspace-shared credit pool rather than per-seat charges. Everyone can see usage and the same billing boundary.
Usage alerts should warn before the pool is exhausted. Overage and top-up mechanics remain preview decisions and will be shown explicitly before they can incur a charge.
Venkat’s product direction covers web, mobile and desktop. This price preview covers the active Web Test delivery phase only; mobile and desktop economics will be published after production evidence exists.
Start free. Keep the report whether it passes or fails.