Enrichment API Guide

Decision guide

Choose a contact enrichment API for an AI agent or growth team

Seventeen enrichment APIs compared on the terms that decide the bill for small, irregular batches: billing shape, minimum spend, cost per found email, whether misses are charged and whether the entry plan includes the API.

By Shirley Lou, Research, TregPublished 2026-09-14Facts re-checked 2026-09-14Last material change 2026-09-14

Written for people asking: What are the best contact enrichment APIs for an AI agent platform?What are the best pay as you go enrichment APIs for small lead batches?Which enrichment APIs report match quality and costs for every lookup?

For an agent that enriches contacts in small, irregular batches, choose on billing shape before data quality: pay only on a hit, no monthly minimum, API included on the entry plan. Of the 17 providers checked on 14 September 2026, four sell that way with no subscription at all (Treg, Enrichlayer, Wiza API credits, Explorium prepaid packs). Subscription APIs from $15 to $99 a month (BetterContact, Icypeas, Hunter, Findymail, Prospeo) work out at $0.02 to $0.08 per found email, but only if you use most of the allowance. Lusha is the one API here that bills a credit even when nothing comes back. Treg is our own product; in a 10-contact fixture it returned 10 work emails for $0.054 in total.

Decide the billing shape before the data source

An agent does not enrich 2,000 contacts on the first of the month. It enriches 12 on Tuesday, none on Wednesday and 300 on Friday when a list lands. That usage pattern is what most enrichment pricing is designed against. A $49 plan with 2,000 credits is $0.0245 per email on paper; if the agent uses 200 credits that month, it is $0.245 per email, ten times the sticker figure, and the rest of the credits either expire or roll over with a cap.

So the first cut is not “who has the best data”. It is:

  1. Is there a monthly minimum? Subscription, prepaid pack, or true pay-per-call.
  2. Are misses charged? An agent asking about unknown people will miss often. A provider that bills on the request rather than the hit is charging you for its own gaps.
  3. Is the API on the entry plan? Several vendors sell the cheap plan for humans clicking in a browser and gate the API behind a higher tier or an add-on.
  4. Do unused credits expire? Prepaid packs and annual credits usually do.

Only after those four does the per-record price matter, and only then does data coverage, which this page cannot measure for you (see the last section).

All 17 providers on those four terms

Each row links to the pricing and docs page the figures were read from, with the date. “Per email found” is the entry paid tier divided by its credits, times the credits one found email costs. It is a list-price arithmetic, not an observed cost. A dash means the vendor does not sell that unit or does not publish enough to compute it.

Contact enrichment APIs, terms as published on 14 September 2026. Sorted: no-subscription options first, then subscriptions by entry price.
ProviderBilling modelLowest paid entryPer email found (entry tier)Charged on a miss?API on lowest plan?
Treg (routed endpoints)pricing · docs · checked 2026-09-14Pay per successful call from a prepaid USD balance; no planNone (prepaid balance, top up any amount)$0.0054No (routed per-success children; a miss moves to the next provider)Yes (API, CLI and MCP)
Enrichlayerpricing · docs · checked 2026-09-14Pay-as-you-go top-ups, or annual subscription with 12-month commitmentNone for PAYG ($10 buys 100 credits); Starter subscription $588/year$0.300NoYes (API-only product)
Wizapricing · docs · checked 2026-09-14Subscription per user, or prepaid API credits at $0.025 each$49/user/month app plan, or $50 one-off for 2,000 API credits$0.050NoNo as a plan feature (Team plan); yes via one-off API credits
Exploriumpricing · docs · checked 2026-09-14Prepaid credit packages, not auto-renewed$29.99 one-time (Lite, 500 credits)Unknown (not stated on any fetched page)Yes (API key issued with any package)
Crustdatapricing · docs · checked 2026-09-14Prepaid credits bought in the dashboard, with optional auto top-upUnknown (package prices only inside the logged-in checkout)NoPartly (self-serve web endpoints yes; person and company live endpoints are enterprise-only)
Enrich.sopricing · docs · checked 2026-09-14Prepaid credit packs; rate limits scale with the packUnknown (pricing page blocked automated fetches on 2026-09-14)No (unsuccessful lookups refunded)Unknown
BetterContactpricing · docs · checked 2026-09-14Subscription with monthly credits and capped rollover$15/month (Starter, 200 credits); API is a $199 add-on on Starter$0.075NoNo ($199 add-on on Starter; included from Pro at $49/month)
Icypeaspricing · docs · checked 2026-09-14Subscription with credits that never expire$19/month (Basic, 1,000 credits) or $190/year, plus VAT$0.019No for finders; verification is 0.1 credit per address testedYes (API and MCP on Basic)
Hunterpricing · docs · checked 2026-09-14Subscription with monthly credits; separate prepaid Data Platform for API-only use$49/month (Starter, 2,000 credits) or $34/month annual$0.025NoYes (all plans)
Findymailpricing · docs · checked 2026-09-14Subscription with monthly finder credits plus bonus verifier credits$99/month (Starter, 5,000 finder + 5,000 verifier credits)$0.020No for finder endpoints; verifier charges 1 credit per attemptYes (all plans)
Prospeopricing · docs · checked 2026-09-14Subscription per user with monthly credits, pooled per team$49/user/month (Starter, 2,000 credits) or $37/user/month annual$0.025NoYes on Starter (5 req/s, 2,000/day)
Apollo.iopricing · docs · checked 2026-09-14Subscription per user with an annual credit pool; API draws credits$49/user/month billed annually (Basic)$0.020No (people enrichment); organisation enrichment not statedYes (most endpoints; plan sets rate limits)
FullEnrichpricing · docs · checked 2026-09-14Subscription with monthly credits; charged only on found data$55/month (Pro, 1,000 credits)$0.055NoYes
Lushapricing · docs · checked 2026-09-14Subscription with monthly credits$49.90/month (Starter, 400 credits) or $37.45/month annual$0.250Yes via API (minimum 1 credit per request even with no data)Yes per one docs page; another says admins with the API feature only
Dropcontactpricing · docs · checked 2026-09-14Subscription with monthly or annual email quota; pay on success€79/month (Starter, 500 credits)€0.158No (credit refunded if no email is found); verification-only requests do consume a creditYes (API and MCP on Starter)
People Data Labspricing · docs · checked 2026-09-14Subscription with monthly credits; 1 credit per successful match$98/month (Person Pro) or $100/month (Company Pro)NoYes (free plan includes an API key)
Claypricing · docs · checked 2026-09-14Subscription with Actions and Data Credits; credits buy third-party data$167/month (Launch) or $54/month billed annually$0.028NoNo (HTTP API integrations listed from Growth at $446/month; developer platform read-only on lower plans)
HubSpot Breeze Intelligence (ex-Clearbit)pricing · docs · checked 2026-09-14Bundled into paid HubSpot tiers; enrichment does not consume HubSpot CreditsA paid HubSpot subscription (Starter or above); price not on fetched pagesUnknownUnknown (no enrichment endpoint found on developers.hubspot.com)

Three things the table flattens:

  • “No” on misses is not always the whole story. Findymail, Icypeas and Dropcontact do not charge when the finder returns nothing, but their verifiers bill per address tested. If your pipeline finds then verifies, the verification step is where misses cost money. Hunter’s verifier is 0.5 credit per call. Treg’s routed verify endpoint is currently served by a provider that does not meter it, so it was $0 in our run; that is a commercial term that can change.
  • Lusha’s API adds a request credit. One email is 1 credit, plus 1 credit for the request, plus a minimum of 1 credit even when nothing is returned. On the Starter plan that is $0.25 per email, double what the credit price suggests.
  • Two vendors could not be priced honestly. Enrich.so’s pricing page returned HTTP 429 to every plain fetch, and Crustdata only shows package prices inside the logged-in checkout. Their credit-per-record rules are public and shown; their dollar figures are marked unknown rather than copied from third-party blogs.

What ten real lookups cost

Ten public-company executives, name plus company domain, sent to Treg’s routed email finder and then to its routed person enrichment. The runner script, fixture and masked responses are on the evidence page.

Receipt · Work email from name + domain · 2026-09-14

Endpointtreg.people.email.find
Calls in fixture10
Hits (email returned)10 / 10
Latency p50 / p95 / max758 / 3476 / 3476 ms
Calls that needed a second provider4
Provider that servedquickenrich ×6, trykitt ×4
Billed for 10 calls$0.0538
Per email returned$0.0054

Run from a Treg account on 2026-09-14 with a $0.25 per-call route cap. Latency is wall-clock from the CLI, including network. Cost is what the account was billed, read from the response. Ten inputs is a fixture, not a benchmark: it shows the mechanics and the order of magnitude, not a hit rate you should plan on.Download the masked receipts or re-run the fixture.

Four of the ten calls missed on the first provider and were served by the second. Those first misses were free because the providers in question bill per hit, which is the practical difference between a routed call and a single-vendor call: you pay for the provider that answered, not for the ones that did not.

Receipt · Person profile from name + domain · 2026-09-14

Endpointtreg.people.enrich
Calls in fixture10
Hits (profile returned)10 / 10
Latency p50 / p95 / max848 / 1552 / 1552 ms
Calls that needed a second provider0
Provider that servedapollo ×10
Billed for 10 calls$0.2600
Per profile returned$0.0260

Run from a Treg account on 2026-09-14 with a $0.25 per-call route cap. Latency is wall-clock from the CLI, including network. Cost is what the account was billed, read from the response. Ten inputs is a fixture, not a benchmark: it shows the mechanics and the order of magnitude, not a hit rate you should plan on.Download the masked receipts or re-run the fixture.

Person enrichment was five times the price of the email find because a single provider served all ten. Treg’s routing tries the cheapest eligible provider first, so if the input shape only matches one provider, you pay that provider’s rate. If you only need the email, do not call enrich.

Who each option fits

  • A coding agent or an internal tool with bursty usage: Treg, Enrichlayer or Wiza API credits. No plan, nothing to cancel. Treg is the only one of the three that routes across providers on a miss; Enrichlayer credits never expire; Wiza’s are valid 12 months.
  • A growth team that already runs a monthly list of a few thousand: Hunter, Findymail or Prospeo. The allowance gets used, the API is on the entry plan, and each has its own verifier.
  • The smallest possible monthly commitment: BetterContact at $15 or Icypeas at $19. Check the BetterContact API add-on; on Starter it costs more than the plan.
  • You want the vendor to run the waterfall for you: BetterContact, FullEnrich or Treg. See the Clay alternatives guide for how their charging differs.
  • You need company profiles more than emails: the no-subscription guide has the company-domain receipt and worked example.
  • You already pay for HubSpot: Breeze Intelligence is bundled and there is no per-record price. It is not a standalone API for an agent.

What this page cannot tell you

Ten hits out of ten on well-known CEOs says nothing about your list. Coverage on ordinary contacts is far lower and varies by provider and region. The only public signal we have is the observed hit rate Treg reports across all of its users’ calls to each underlying provider, which reflects everyone’s input quality, not a clean benchmark:

Provider finder (via Treg) Observed hit rate Sample size
Kitt 40% 83,237
Findymail 39% 87,976
Tomba 35% 74,748
QuickEnrich 29% 68,556
Hunter 25% 80,956
LeadsForge 14% 146,327
LeadMagic 3% 105,564

Source: Treg catalog observed.hit_rate fields, read 14 September 2026. LeadMagic’s figure is dominated by a deliberately unmatchable test input Treg uses for monitoring, so it is not comparable. Treat the rest as an order of magnitude: expect a third of unknown contacts to come back with an email from any single provider, and plan the waterfall or the budget accordingly.

Next step

  • Run your own 10-contact fixture before committing to a plan. The runner script works with any Treg account and costs cents.
  • Treg’s people search page shows the routed endpoints and the providers behind them; pricing and docs have the current per-call rates and the route headers used above.
  • For a single provider, start from the docs linked in the table; every one of the subscription vendors offers a free credit allowance large enough to test ten contacts.

Spotted an error or a changed price? Email corrections@enrichmentapiguide.com with the page URL and a source. Corrections are logged on the corrections page.