Price the clean-up of a CRM, a supplier master or a stewardship queue before you take it to budget.
You own a company master you have to defend, and sooner or later somebody asks what fixing it costs. The whole answer is on this page: what one decision costs, what the four plans hold, and what a 100,000-row file comes to. We bill a resolved company — never a row, a lookup or a seat — and a record we refuse to call costs you nothing. The published plans are Free at $0 a month with 300 credits, Starter at $49 a month with 1,000 credits, Growth at $199 a month with 5,000 credits and Scale at $799 a month with 25,000 credits; paid subscriptions are purchased in workspace Billing through Stripe, with VAT added where applicable and no automatic usage or overage charges.
Self-serviceCreate your workspace online and verify your email to start matching company records. Free starts with 300 credits and no card. Paid monthly plans are available in workspace Billing; VAT is added where applicable. Usage is tracked without automatic overage charges. Trust and security
billing unit
A resolved company
monthly plans
$0–$799
cheapest paid credit
$0.032
payment method
Stripe Checkout
billing statusPaid subscriptions live
planned free evaluation
1,000 one-time credits at full research depth, 30 days, no card. The full-pipeline evaluation is arranged separately and is not activated by signup. Request it for your workspace; we confirm availability, sources and the start date before activation.
The same rule applies over HTTP and over MCP, on every plan. The prices are defined in the product’s own code: this rule is read off the same module that writes the usage ledger and serves GET /v1/reference/pricing, so you can check what is configured.
the credit rule, in full
1credit
A decision reached by the deterministic or probabilistic stage. CPU only, no egress.
5credits
A decision that needed LLM adjudication or agentic research.
0credits
An abstention, at any stage, including the ones that spent the most getting there.
0credits
A byte-identical replay. Per tenant, per exact content, and with no expiry.
0credits
An input that normalised to no identifying signal. There was nothing to answer.
what a credit buys
You pay for a decision, never for an attempt. A decision the identifier rules or the scoring can settle costs 1 credit: it never reaches a model and never leaves the machine. Adjudication and live registry research cost real money to run, so a decision needing either costs 5. How often a row needs the deep stages is counted on one fixture below.
why the abstention is free
A record we cannot support at the 0.90 confidence floor comes back unresolved, with a ranked shortlist, a typed reason, the column that would have decided it and a review id. It costs nothing at every stage, including the ones that spent real money on adjudication and research getting there.
Charging for it would corrupt the incentive that makes it trustworthy. If a refusal earned revenue, the cheapest way to raise revenue would be to refuse more.
It is an acceptance contract, not calibration evidence: it says what we refuse to return, not how often an accepted match is right. Read it with the coverage it costs.
two controls on the spend
Free abstentions put the expensive tail of the cascade on our side of the ledger. Two controls bound it.
control one
A depth cap on the call
Set the depth to probabilistic and the row cannot cost more than 1 credit. It caps egress as well as models: the registry candidate pass and the enrichment gather are both gated by it, so a capped call reaches no external service at all.
The quota bounds research spend. There is no consolation fee for an abstention. Exhausting the quota degrades the call to shallow depth and returns an unresolved; it never stops a batch with a payment error.
The allowance is the tenant’s, not the key’s. Three keys draw on one allowance, and rotating a key does not hand you a fresh one mid-period.
Workload planning
Calculate the credits for your scenario
Use your own assumed outcome counts to compare the published plan allowances.
Enter a hypothetical mix of outcomes, or import the aggregate scenario CSV. Your counts stay in this browser. This tool calculates credits and plan allowances. It does not inspect company identities or forecast how your file will resolve.
The deep unresolved count is part of the unresolved total. It costs zero credits and still consumes a deep attempt. Exact replays and empty inputs are separate rows.
Compare the scenario with the published monthly allowances
Plan
Monthly subscription
Credits included
Deep attempts included
Scenario fits both
Free
$0/month
300
0
Exceeds allowance
Starter
$49/month
1,000
500
Exceeds allowance
Growth
$199/month
5,000
3,000
Yes
Scale
$799/month
25,000
20,000
Yes
Paid subscriptions are purchased in workspace Billing. VAT is added where applicable. The table is an allowance comparison, not a usage quote. There are no automatic usage or overage charges. A real run can produce a different mix of outcomes.
published monthly plans
Choose a plan when you know your workload
The 300 credits Free plan is capped at probabilistic depth. Each card prints what a credit and a deep answer cost on that plan, so you can check the ladder yourself.
Free
$0a month
Credits
300 credits
Deep attempts
No deep resolutions
Depth ceiling
Capped at probabilistic
A credit costs
No charge
A deep answer costs
Not reachable
A hard no-egress cap: no model provider and no registry either. You get the fast, deterministic two-thirds of the cascade against your own submission, which is genuinely limited and worth knowing before you build on it.
Separate from the planned 1,000-credit evaluation below, which is one-time and runs at full research depth. Every Entity Bind workspace opens on this plan.
The whole pipeline, including adjudication and research, on a file you are still sizing.
A row that reaches the deep stages costs 5 credits whether it took one adjudication or a full research pass, so an unfamiliar file spends this allowance faster than its row count suggests.
A standing integration against a CRM or a supplier master.
Sized for a standing integration, not a single file: a nightly CRM sync, a supplier onboarding queue, an agent checking a counterparty. Replays make the second night much cheaper than the first.
25,000 included credits, three times the rate limit, and the only plan carrying resale rights on the open-licensed subset — with storage permission and caller resale rights checked separately for each fact. Available in workspace Billing at $799 per month, plus applicable VAT.
Storage permission and caller resale rights are checked separately for each fact. Resale covers the open-licensed subset; discuss the applicable licence terms before building a redistribution product.
Credits run out before deep attempts do on every paid plan. Charging for each attempt in the deep quota would take Starter, 2,500 credits against the 1,000 included; Growth, 15,000 credits against the 5,000 included; Scale, 100,000 credits against the 25,000 included. Abstentions cost no credits and still consume attempts, so an ambiguous file can exhaust its quota first.
when a quota runs out
An enforced deep quota caps later calls at shallow depth, and cases needing more work abstain. Quotas are not enforced while usage metering runs in shadow mode; the planned evaluation carries its own limits, set out below.
Start on Free to inspect your own rows. Choose a paid plan in workspace Billing when you know your workload; contact us if your file needs more than the published allowances.
Create your Free workspace online, with 300 credits and no card. Workspace owners purchase paid monthly subscriptions through Stripe Checkout, with VAT added where applicable. Usage credits are tracked separately without automatic overage charges.
live today
Create and verify your workspace, upload a CSV, map your columns and inspect its results and decision receipts. Free supports shallow matching. Create your own API keys for integrations; activate registry checks through a paid subscription.
separate evaluation
1,000-credit full-pipeline evaluation. The full-pipeline evaluation is arranged separately and is not activated by signup. Request it for your workspace; we confirm availability, sources and the start date before activation.
Review the plan, recurring total and payment terms before purchase. Signup does not activate the separate full-pipeline evaluation or start its clock.
worked arithmetic
What a 100,000-row file costs
What a file costs is set by the columns it carries, not by its row count. Here is the same 100,000 rows in three compositions, with every intermediate figure printed.
the escalation rates this uses
How often a row leaves the cheap half of the cascade, counted on the continuous-integration fixture: 705 records over 250 UK-registered companies, run on 2026-09-05, of which 499 resolved. The three rows below are the whole fixture.
What you send
What the row carries
Left the shallow stages
Reading
Identifier
A company number or an LEI in a column of its own
0/250
All 250 were settled by the identifier rules. None reached a model.
Name and website
A company name with a domain beside it
1/250
249 of 250 bound at the shallow stages. The one that did not had a second near-automatic candidate, and the guard withheld the merge.
Name only
A typed company name and nothing else
205/205
All 205 left the shallow stages. The deep stages on this fixture are offline stubs, so all 205 abstained and cost nothing — the floor of a live run, not a forecast of it.
End-to-end resolution on the internal gated acquisition fixture with offline stub models. All 499 accepted matches were correct in this fixture. The current 99% product-target diagnostic fails; this run does not measure live model conversion or independently establish calibration.
A supplier master out of an ERP that has carried a registration-number field since it was configured.
Row by row
Shape
Rows
Deep
Floor
Ceiling
Identifier
80,000
0
80,000
80,000
Name and website
15,000
60
14,940
15,240
Name only
5,000
5,000
0
25,000
Total
100,000
5,060
94,940
120,240
94,940 – 120,240 credits · at $0.032 each, the Scale rate
$3,034 – $3,843
Capping the depth at probabilistic pins this file to the floor, 94,940 credits.
Mixed
A CRM account table after five years of self-service sign-ups and one acquisition.
Row by row
Shape
Rows
Deep
Floor
Ceiling
Identifier
30,000
0
30,000
30,000
Name and website
40,000
160
39,840
40,640
Name only
30,000
30,000
0
150,000
Total
100,000
30,160
69,840
220,640
69,840 – 220,640 credits · at $0.032 each, the Scale rate
$2,232 – $7,052
Capping the depth at probabilistic pins this file to the floor, 69,840 credits.
Name-only
One column of typed company names: a conference list, a prospect sheet, a free-text employer field.
Row by row
Shape
Rows
Deep
Floor
Ceiling
Identifier
0
0
0
0
Name and website
0
0
0
0
Name only
100,000
100,000
0
500,000
Total
100,000
100,000
0
500,000
0 – 500,000 credits · at $0.032 each, the Scale rate
$0 – $15,980
Every row here leaves the shallow stages, so a depth cap costs nothing and resolves nothing.
how to read the band
The floor assumes every escalated row abstains, which costs nothing and resolves nothing. The ceiling assumes every one is decided at the deep stages, at 5 credits each. A real file lands between them, and the product’s own pre-flight estimator draws the same band — zero to 5 credits a row — refusing with a stated reason when it cannot tier a row.
the limits of this arithmetic
Not a quote. There is no overage rate and no top-up, and 100,000 credits is 4 times the largest published allowance, so a file this size is a conversation — which is what the Scale plan is for. The escalation rates come from one internal fixture over one register with its deep stages stubbed out; your columns and your registers are not that fixture.
Internal fixtures only. The historical 2026-08-29 per-tier recording covers 317 GB companies with identifier-rich arrivals seeding the catalogue first; the context tier includes country and registered-town prose. Its legacy replay has no full-prompt-verified calls and cannot certify current calibration. The September 5 CI reports cover 1,176 labelled pairs and 705 records over 250 companies using offline stubs. These are not independent customer holdouts or live production validation; the 99% minimum-input product target remains unmet. Scope is company and organisation identity only — not people, vessels, bank-account ownership, sanctions or emissions. Results may not be used for eligibility decisions about individuals, including employment, credit and tenancy.
The composition of your file is the one input this arithmetic cannot guess. Upload your CSV in your Free workspace to inspect your columns, or contact us for an assisted composition preview and a scoped Sprint.
Resolution is bound to a content fingerprint: the SHA-256 of the record you submitted, with the server-minted id excluded. Resubmit the same bytes next month or next year and the same binding comes back at no cost. There is no expiry.
It is free on every plan, including the free one, and there is no premium replay tier. A reviewer’s verdict writes a fingerprint binding too, so a record settled in the queue replays without re-running the cascade.
Two things a replay is not. It does not freeze the golden record — you get the entity as it stands now, which may carry facts it did not carry last month. And it is not a way past the confidence floor: a stored confidence that has fallen below 0.90 is withheld and replayed as unresolved.
caveat one
It is per tenant
The fingerprint table is keyed on the fingerprint and the tenant together, so one customer’s spend never subsidises another’s.
caveat two
It is per exact content
One changed character is a fresh chargeable resolution: a corrected postcode, an added phone number, a trailing space. Erasing the entity drops the binding with it, so those bytes are chargeable again.
the planned free evaluation
1,000 credits, and room for the difficult cases
Give your agent room to test your companies, inspect the citations and compare the results. Deterministic matching, adjudication and registry research, with the same evidence and confidence rules as the paid product.
credits buy decisions
Up to 200 deep answers, or up to 1,000 shallow ones, or a mix. Allowance limits, not a promise that every company will resolve.
deep attempts are counted separately
1,000 of them. An abstention uses no credits and still uses an attempt, so bring the same-name companies and the incomplete records.
reading costs nothing
Entity, receipt and reference reads use no credits, and neither do exact replays.
1,000 one-time credits, valid for 30 days from your first resolution. No card required. Full resolution depth, with configured sources and their existing licence terms.
The planned allowance includes up to 1,000 deep attempts, including those that abstain. Credits and attempts are separate limits. Provider-spend protection may pause work explicitly; unused credits stay visible and the pipeline is never silently downgraded.
One illustrative outcome mix, not a predicted resolution rate
Outcome
Cases
Credits
Decisions settled by matching alone (shallow)
300
300
Decisions needing adjudication or research (deep)
120
600
Cases with too little evidence, left unresolved
100
0
Example total
520
900
100 credits remain in this example. What a run costs depends on the stages it reaches.
The full-pipeline evaluation is arranged separately and is not activated by signup. Request it for your workspace; we confirm availability, sources and the start date before activation.
the commercial arrangement
5 things the price list does not carry
2 of them are decisions: pricing is per account, and one tenant's spend never subsidises another's. Of the rest, 3 are mechanisms we have not built. Every kind is on the page so that nothing here changes shape at the point of purchase.
No overage
not built
There is no rate for a credit past your allowance, because there is no mechanism to charge one.
No top-ups
not built
You cannot buy credits mid-month. A file that needs more than a plan holds is a conversation about the file.
No annual discount, yet
not built
An annual price at roughly 17% off is proposed in the commercial model and has not been built.
No per-seat pricing
a decision
Pricing is per account and per resolved company. Another colleague costs nothing, and so does another key.
No cross-customer discount
a decision
A company another tenant has resolved is full price for you. The fingerprint is keyed per tenant, which is a privacy property before it is a pricing one.
Create a Free workspace online and purchase a paid subscription in workspace Billing. The meter, ledger and published credit rule show your usage. For an assisted Merge Safety Sprint or a procurement arrangement beyond this page, contact us and we will tell you whether it exists.
Free signup and the published paid plans are self-service. Contact us for an arrangement outside those plans; a person will tell you whether it exists and what it would cost.
Teams with a company master they have to defend — an MDM or stewardship queue, a CRM after a migration or a merger, a supplier file before a payment run, counterparty onboarding that has to survive an audit. You are pricing work somebody will ask you to justify, so every figure on this page is one you can hand them.
wrong for
Buying breadth as the product: we hold a fraction of an incumbent’s records and are not building a global company graph. Entity Bind resolves companies and organisations only. It does not resolve people, vessels, bank-account ownership, sanctions exposure, emissions or geolocation, and its results may not be used for eligibility decisions about individuals, including employment, credit and tenancy.
The retention, erasure and tenancy detail a security review asks for is on the trust page, with where assurance stands: no SOC 2, no external validation of the accuracy figures, no customer holdout. Trust and security →
Questions about cost
What people ask before they send anything.
How much does a 100,000-row file cost?
It depends on the columns rather than the row count, and you can work it out yourself. A row decided by identifier rules or probabilistic scoring is 1 credit; a row that needed adjudication or research is 5; an abstention, a byte-identical replay and a row that normalised to no identifying signal are all 0. So 100,000 rows all carrying company numbers is 100,000 credits, and the same file where every row went deep is 500,000. At the Scale rate of $0.032 a credit, the cheapest paid rate we publish, that is roughly $3,196 and $15,980. That is more credits than any priced plan includes and there is no overage rate, so a file this size is a Scale conversation, quoted against your volume. Before any quote, we run a free preview over your file. It makes no model or registry calls, tiers your rows and reports composition, never a predicted resolution rate. Then we talk it through. Start free, then choose a paid plan in Billing.
Are you charging cards today?
Yes, for paid monthly subscriptions purchased through Stripe in workspace Billing. Review the recurring price, applicable VAT and payment terms before purchase; Stripe provides the subscription invoices. Free needs no card. Usage metering is separate: calls are recorded in the ledger, and usage creates no automatic usage or overage charge. The usage block states whether quota enforcement is enabled for the deployment; an `enforced: false` response means the call is metered without enforcing the usage quota, not that a paid subscription is free. The plan prices are defined in code and served from the same constants that meter, so the number in a response and the number on the page cannot drift apart.
Why is an abstention free?
An unresolved identity costs zero credits even when adjudication or research did the work. Deep attempts are tracked separately, because those calls still carry a provider cost. Existing monthly plans can reduce depth when their attempt quota runs out. The planned 1,000-credit evaluation instead requires an explicit limit response or a disclosed pause when research cannot continue, so a budget limit cannot be mistaken for an evidence-based abstention. That change is not live yet.
Do we pay twice for the same record?
No. Resolution is bound to a content fingerprint — the SHA-256 of the submitted record with the server-minted id excluded — so resubmitting the same bytes returns the same binding at zero credits, on every plan, with no expiry. It is not a cache with a 24-hour window, which would be a promise about our infrastructure that you cannot verify; it is a property of your data. Two caveats travel with that claim. It is per tenant, so there is no cross-customer discount for a company somebody else has already resolved. And it is per exact content: one corrected postcode, one added phone number or one trailing space is a different fingerprint and a fresh resolution.
Can I resell the data?
On the open-licensed subset, on the Scale plan, and with attribution. It is enforced by the write path rather than by a clause: every fact carries a licence tier, and the repository raises before any write on a pass-through fact, so a violation cannot even partially persist, and the field-by-field assembly of the golden record drops pass-through material before the record is built. Resale rights are published per field — `licence_permits_resale` is a property of the data's own licence, `resellable` is that plus the right your plan bought — and an unreadable licence reads as not-resellable rather than as permitted. Companies House is OGL v3.0 and SIRENE is Licence Ouverte 2.0, and both require attribution. Talk to us before you build a redistribution product on it.
What does the free plan actually do?
The existing Free plan includes 300 credits a month, no deep resolutions, and a probabilistic depth cap. It makes no model or registry calls, so it is useful for wiring an integration but does not test research coverage. Every Entity Bind workspace starts on it, and a CSV upload in Matching mode runs there. The planned 1,000-credit evaluation is a separate one-time offer with full research depth. Request access to test that pipeline on your own companies; a sample file is optional. The full-pipeline evaluation is arranged separately and is not activated by signup. Request it for your workspace; we confirm availability, sources and the start date before activation.
Why 5 credits for a deep resolution rather than 2 or 10?
Because it is a price rather than a measured cost ratio, and the code comment says so. The measured input is narrower: one LLM adjudication was measured at £0.0240 over 243 live calls, and the cost model carries $0.0293 per adjudication verdict; at about 1.73 verdicts per escalated record, that is about $0.051 for a record that escalates. Per thousand rows it is about $0 for an identifier-rich file and about $50 for one that is all names, because escalation is the whole cost: on the gated acquisition fixture 0 of 250 identifier-bearing records and 1 of 250 digital-tier records escalated, against 205 of 205 name-only ones. The 5:1 multiple is charged identically whether the answer came from one adjudication or a full research pass, and research costs several times more. Shadow-mode metering exists to check that ratio against reality before anybody is billed on it.
How does the pricing shape compare with an incumbent's?
Badly today, better over time, and the reason is structural. An incumbent's marginal cost per lookup is a database read, a hundredth of a cent, because the database is already built. Ours is per-query inference on the records that escalate. What differs is the direction of travel: a company already held in the catalogue with its identifiers binds deterministically for free, so our marginal cost falls as the catalogue fills, while theirs is already at the floor. That is a reason to send a file rather than a reason to believe a projection, and it is why we cap depth per record and publish the credit rule instead of quoting a blended per-record price.
the ask
Start with your own company records.
Create your Free workspace with 300 credits and no card. Upload a CSV to inspect your own rows, then choose a paid monthly plan in Billing when you need more depth or capacity. We write nothing to your systems.