Pricing Projection

Each buyer pays for the evidence it needs.

These are the business plan's planned list prices for 2028–2030. Nobody is paying them today. They will be tested with design partners and changed by what buyers actually pay.

Interactive · configurator

Build an indicative quote.

Pick a buyer and a scope. The estimate stays inside the plan's projected range for that SKU.

Who is buying

AI systems in scope 1

Chatbots, copilots or agents tested as separate targets.

Testing cadence

Self-hosted / VPC runner (transcripts never leave your environment) GRC / TPRM integration and SSO Agent testing through the dry-run tool-call proxy 2027 design-partner pricing (where offered)

FI Compliance Assessment

$20k

indicative annual list price · projection, not a quote

All SKUs Projection

The planned price list.

SKU	Buyer	List price (2028–2030)	Includes
Vendor Attestation	AI vendors serving FIs	$55–80k	Continuous testing in vendor CI; shareable attestation; FI viewer free. Design-partner pricing $25–40k in 2027.
FI Compliance Assessment	Banks and CUs $1–10B (compliance owner)	$20–27k	Annual assessment + quarterly retest of the FI's configuration; Tier 1 policy-as-code checks; viewer on vendor attestations; evidence pack
Partner-delivered	Institutions under $1B via pen-test/audit firms	~$9–11k net	The partner's annual AI assessment runs on Countersign; net of a 50% partner share
Fintech	Fintechs and lenders	$40–60k	CI integration, API, conduct and security suites
Regional	Banks $10–100B	$90–145k	Multiple systems, GRC integration, SSO, self-hosted runner option
Enterprise	Banks over $100B	$250–420k	Agents via dry-run proxy, VPC/on-prem, custom suites, platform-finding ingestion
Adjacent regulated	Insurers, broker-dealers, payers	$60–78k	Same engine, industry policy packs (from 2029)

Expert services are planned to stay at about 10% of revenue. Agent testing is priced separately because setup is real work; no price has been set.

Where revenue would come from Projection

Banks are the credibility customer, not the revenue base.

Base-case 2034 ARR by segment. U.S. banks above $10B are only about 155 institutions, so the model leans on AI vendors, fintechs and, later, other regulated industries. Retention is modeled by segment: community banks and credit unions are assumed near 100–101% net revenue retention, regional and large banks about 119%.

Segment	Logo churn	Expansion	NRR
AI vendors	10%	30%	117%
Fintechs	12%	30%	114%
Banks/CUs $1–10B	8%	13%	101%
Partner-delivered	12%	14%	100%
Regional banks	5%	27%	119%
Large banks	4%	26%	119%
Adjacent regulated	8%	25%	114%

Steady-state assumptions from model.json, including 2–4% downsell; expansion ramps up over the first three years.

Pricing questions

Why it is priced this way.

Why does the vendor pay, not the bank?

For vendor-hosted chatbots, the vendor ships the changes, owns the test environment and is the party that can authorize testing. One attestation serves all of a vendor's FI customers, who view it free and buy only the checks of their own configuration.

Why an annual assessment rather than a subscription for community banks?

Banks already buy an annual assessment the examiner can see. The entry SKU matches that: an annual assessment with quarterly retests. Continuous testing is the upsell.

Why is the community-bank price so low?

Because that is what the compliance-testing budget at a $1–10B institution supports, and many credit unions have signing thresholds below it. The plan's own unit economics show this segment is only worth serving through vendors and partners.

Does a higher plan raise our own LLM bill?

Yes. Statistically valid testing means many trials, which costs target-side inference. The plan commits to publishing target-side call budgets per plan.