How Per-Report Credit API Pricing Works (and How to Forecast It)
Per-report pricing looks simple until the first invoice. Here's what actually gets billed, the costs that hide in credit data contracts, and a forecasting model you can drop into a spreadsheet, with worked examples at three volumes.
On this page
- What you’re paying for with a per-report price
- The three credit report API pricing models
- What counts as a billable report
- The forecasting model
- Monthly or yearly: the breakeven is three months
- Turn cost per report into cost per outcome
- Controls that keep the invoice predictable
- Costs that aren’t on any pricing page
Most credit report API pricing has two parts: a platform fee (monthly or yearly) and a charge for each report you pull. Your bill is the platform fee plus the number of billable reports times the per-report price, plus any add-ons. The hard part isn’t the formula. It’s knowing which requests count as billable, and estimating how many you’ll make before the month starts.
Both are solvable. Below is what’s actually being priced, the costs that never show up on a pricing page, and a forecasting model you can build in ten minutes, worked through at three volumes with real prices.
What you’re paying for with a per-report price
A credit report isn’t free for anyone in the chain. The bureau charges for the data, the score developer charges for the score, and whoever sits between you and the bureau charges for access, compliance oversight and the API itself. Per-report pricing bundles all of that into one number you pay only when you use it.
That’s why the model fits most businesses that pull credit as part of a sale or an application. Your costs rise with your volume, and a slow month costs less than a busy one.
The alternative, going direct to a bureau, usually looks very different on paper. It typically means a membership application, a site inspection of your office, security audits, setup or membership fees and monthly minimums, before you see your first report. For high-volume lenders, that can make sense. For a dealership, a property manager or a 12-person brokerage, the fixed costs and the weeks of paperwork usually don’t.
The three credit report API pricing models
You’ll run into three basic structures when you shop for credit data.
Pure pay-as-you-go. No platform fee, a higher price per report. Good for very low or unpredictable volume. The catch is that per-report prices are usually highest here.
Platform fee plus per report. A monthly or yearly subscription that covers API access, keys and tooling, plus a flat per-report rate. This is how our pricing works: $49 a month or $149 a year, plus $7.50 per report, with no cap on monthly volume. The platform fee is small enough that it rarely decides anything. The per-report rate is what you’re really forecasting.
Committed volume. A negotiated per-report rate, often in exchange for a monthly minimum or annual commitment. Lower unit price, but with a minimum you pay for the commitment whether you use it or not. With us, there is no monthly cap, and volume rates are the conversation to have once you reach 1,000 reports a month: email [email protected].
If you’re still comparing providers on features rather than price, our credit score API buyer’s guide covers what to check before you integrate.
What counts as a billable report
This is where invoices surprise people. Ask every provider these questions, in writing.
Is a no-match free? If you send a phone number and nobody’s found, do you pay? With us, no match means no charge. Some providers bill every request.
What about frozen or unscorable files? If the person is found but their file is frozen or too new to score, most providers bill it, because the bureau still did the work. We do too, and the response tells you which case it was.
Are retries billed twice? Networks fail. If your code retries a request that actually succeeded, you can pay twice for the same person. Our API accepts an Idempotency-Key header so a retry never charges twice. If a provider doesn’t support something like that, build your own de-duplication.
Are test calls free? Test keys (ours start with pncs_test_) return example people and are never charged. Make sure your QA environment can’t accidentally use live keys.
What do add-ons cost? Full account-by-account history is usually extra. Ours is $2 more per report for tradelines, for $9.50 total, and you can add it at request time or later to a lookup you’ve already run. If you’re not sure you need that detail, read what a tradeline is first; for many decisions the score and credit profile are enough.
Do lookups through an AI assistant cost more? Not with us. Our MCP server for AI assistants comes with the API plan, and a lookup an assistant runs is priced exactly like one your code runs.
The forecasting model
The whole model fits in three lines:
Monthly cost = Platform fee
+ (Requests × Match rate × $7.50)
+ (Billable reports × Tradeline share × $2.00)
Three inputs drive it.
Requests. How many lookups you’ll send. Start from your funnel, not a guess: how many applications, consented leads or bookings do you get a month, and what share of them will you check? Remember that every request needs a permissible purpose, usually the applicant’s written consent or an application they started, so your request count can’t be bigger than your consented pool. The list of valid purposes is in 15 U.S.C. § 1681b. The permissible purpose guide explains why.
Match rate. The share of requests that find a person. It depends heavily on your data: real mobile numbers from applicants match far more often than old landlines or typo-heavy web forms. Don’t borrow anyone else’s number for this. Run your first few hundred live requests and measure it.
Tradeline share. The share of billable reports where you’ll add the full account history. Many teams add it only for approvals above a certain amount, or for files near a decision boundary.
Worked example 1: a small loan brokerage
A two-person brokerage gets about 140 consented leads a month and checks 120 of them. Say their measured match rate is 85%, and they add tradelines on 1 in 5 billable reports.
| Line | Math | Cost |
|---|---|---|
| Monthly plan | $49.00 | |
| Billable reports | 120 × 0.85 = 102 × $7.50 | $765.00 |
| Tradelines | 102 × 0.20 ≈ 20 × $2.00 | $40.00 |
| Total | $854.00 |
On the yearly plan, the platform fee drops to about $12.42 a month ($149 ÷ 12), bringing the total to about $817.
Worked example 2: a multi-rooftop dealer group
Three stores run soft-pull pre-qualification on every customer who agrees at the desk or online: about 600 requests a month. Their match rate is higher, say 90%, because these are customers’ own mobile numbers. They add tradelines on half the reports to see current auto payments.
| Line | Math | Cost |
|---|---|---|
| Monthly plan | $49.00 | |
| Billable reports | 600 × 0.90 = 540 × $7.50 | $4,050.00 |
| Tradelines | 540 × 0.50 = 270 × $2.00 | $540.00 |
| Total | $4,639.00 |
Worked example 3: crossing the line
A property management company with 1,300 consented applications a month at a 90% match rate would hit about 1,170 billable reports. That’s past the 1,000-a-month mark where an account qualifies for volume rates, so it’s the point to talk to us about lower per-report pricing rather than spreading requests across accounts.
Monthly or yearly: the breakeven is three months
The plan choice is the easiest money in this whole post. Twelve months on the $49 plan is $588. The yearly plan is $149, which saves $439.
Flip it around: the yearly plan costs less than four months of monthly. So monthly only wins if you’re sure you’ll use the API for three months or fewer ($147). If you’re building a real integration, pay yearly.
Turn cost per report into cost per outcome
$7.50 a report means nothing until you tie it to revenue. The useful number is cost per closed deal.
Say a broker’s data shows that 1 in 8 checked leads eventually funds. That’s 8 × $7.50 = $60 of credit data per funded loan, before tradelines. If the average commission is $1,500, credit data is 4% of revenue on those deals, and it likely saved hours of calls to leads who couldn’t qualify. (These are illustrative numbers. Use your own conversion rate.)
The same lens tells you where not to spend. If you’re checking leads that never pick up the phone, move the lookup later in the funnel. If you’re adding tradelines to every report but only using them on 1 in 10, add them only when needed, since you can attach them to a past lookup afterward. The loan broker lead scoring post goes deeper on where in the funnel to check.
Controls that keep the invoice predictable
A forecast is only useful if the actual bill tracks it. Build in these guardrails from day one.
- Spending limits. Cap monthly spend at the account level so a bug or a runaway script can’t surprise you.
- Separate keys per system. Your CRM, your website form and your internal tools each get their own key. Usage logs then show exactly which one drove the bill, and you can turn one off without breaking the others.
- IP allow lists. Only your servers can use your live keys.
- Webhooks instead of polling. Let results push to you rather than re-requesting. Our guide on routing leads with webhooks shows how.
- A usage check in your dashboard.
GET /v1/usagereturns lookups used this month. Put it on a screen your ops lead actually looks at. - Seasonality buffer. Tax-refund season, spring moving season and model-year sales all spike volume for different industries. Forecast your peak month, not your average, and add 15 to 20% headroom.
Costs that aren’t on any pricing page
The per-report fee is the visible cost. A few others are worth putting in the spreadsheet.
Data protection. Credit data is sensitive. Many financial businesses are covered by the FTC’s Safeguards Rule, 16 CFR Part 314, which requires a written information security program. Storing less (keep the decision and the score, not a copy of every report forever) lowers both risk and cost.
Adverse action workflow. If you deny or change terms based on a report, the FCRA requires a notice naming the bureau. Our API returns the score’s key factors in every response, which makes those notices easier to generate, but someone still has to build the template.
Integration time. One request in, one JSON answer out is usually an afternoon of developer time, not a quarter. Still, budget for it. The CRM integration guide shows a typical setup.
If you want to see the full list of what comes back for each price, the API page has the report contents and endpoint reference, and the FAQ covers billing questions. For background on which companies actually hold consumer credit files, the CFPB keeps a list of consumer reporting companies.
And if your volume is headed past a few hundred a month, read bulk credit checks for a list of phone numbers before you design the batch job.
