studio
€390
one-time
A single message round on one audience scope — segment-level reaction distributions, a printable watermarked report, and a downloadable trail of every model call behind it.
Buy via StripePricing
One-time pilot packages
No subscription during the pilot phase — buy a package, we run it with you hands-on, and you keep the reports, the audit trail and the caveats.
team
€990
one-time
A variant-comparison round (2–5 variants on the identical audience). The ranking comes with a margin verdict: a winner is named only when the lead survives the wording effect measured on that run.
Buy via Stripeagency
€2.400
one-time
A multi-market round: each country pack runs in its own language, then the results are combined by population weight — that arithmetic happens in code, never delegated to the model. Coverage is reported, and a region below 70% is flagged as partial.
Buy via StripeIn every package
- Segment dispersion is reported separately from the headline number. If the segments barely differ, the result says so instead of hiding it.
- No winner is declared unless the lead is wider than the wording effect measured on that run — otherwise you get “no clear winner”, which is the honest answer.
- A downloadable audit trail of every model call, and a SYNTHETIC SIMULATION watermark that cannot be switched off.
Enterprise
Workspace-specific calibration, custom taxonomies, volume runs and API access are scoped per engagement. Tell us what you are testing and we will quote it.
[email protected]Prices are pilot-phase list prices (hypothesis, may be revised). Payment via Stripe; delivery is hands-on during the pilot — we run your tests with you and iterate until the report is useful. Political and election use is excluded by the terms of service.