Calibrated confidence for every address. Even the catch-alls.
No credit card. 100 free verifications — credits never expire.
{
"status": "deliverable",
"confidence": 0.96
}Example response — verify a real address above.
3 free verifications/day, no signup required
Recognizes 29 mailbox providers and 7 SEGs. See the list ↓
How we measure accuracy.
curl -s https://api.skink.dev/v1/verify \
-H "Authorization: Bearer $SKINK_API_KEY" \
-H "Content-Type: application/json" \
-d '{"email":"jane.doe@acme.com"}'{
"email": "jane.doe@acme.com",
"status": "deliverable",
"confidence": 0.91,
"reasons": [
"Address resolved with high confidence",
"Calibrated against real delivery outcomes"
],
"recommendation": "send"
}No SDK required — plain HTTPS in, JSON out. Node, Python, and Ruby SDKs are available if you’d rather.
share of B2B addresses on catch-all domains, no clean answer possible — MailerCheck's own 33,000-list study
top score among 15 verifiers tested — Hunter's own 3,000-email benchmark of itself vs. competitors
false-positive rate on unresolved addresses, 30% of real contacts missed — Allegrow's own 1,222-email test, ZeroBounce standard
false-confirm rate on fake mailboxes in our own controlled catch-all test — skink's own ground-truth test — see /accuracy
Why it wins.
Protected inboxes, still resolved
Security gateways silently return unknown
Attempts resolution behind SEGs where conventional verification returns unknown.
Exposed reasoning
No explanation for why an address was marked risky or invalid
reasons[] on every result — not just a code.
Calibration that never stops
Static thresholds that never adapt to your actual bounce data
Re-checked nightly against real delivery outcomes.
Honest, not manufactured
Some verifiers proxy-hop to force an answer out of a stalling server.
No forced proxies — a real risky beats a fake deliverable.
Parked isn't always dead
Most verifiers blanket-flag any parked domain as undeliverable.
We check whether it's forwarding before we call it invalid.
How we compare.
Checked against public vendor docs, July 2026 — not an independent audit. See /accuracy for the full competitor matrix.
Providers we support.
Buy a pack, use it whenever.
1 credit = 1 verification. Unknown results are free. SMTP checks and catch-all resolution included at every tier — no paid add-on.
100 credits, no card required — enough to see real accuracy on your own list before you spend anything.
10,000 credits
$0.0049 / email
Your first real batch. Validate a list before a campaign goes out.
Buy Growth50,000 credits
$0.0039 / email · 20% cheaper than Growth
Monthly list hygiene on autopilot.
Buy Scale200,000 credits
$0.003 / email · 39% cheaper than Growth
Built for teams running cold outreach at real throughput.
Buy Volume500,000 credits
$0.0026 / email · 47% cheaper than Growth
Established sending programs — half a million credits without jumping straight to Max.
Buy Business1,000,000 credits
$0.002 / email · 59% cheaper than Growth
A million credits — our deepest per-credit rate, on the same shared infrastructure as every other tier.
Buy MaxRunning past a few million verifications a month, or have other custom requirements? Talk to us before you buy another Max pack.
Save a card once. No subscription.
Your pack repurchases automatically.
Same price as buying it by hand.
One click in Billing.
Your email list is never our product.
We do not sell or rent customer contact lists, use them for advertising, or build a contact database from them.
Real-time requests can be processed without retained recipient-level history.
Choose how verification history is stored and retained.
Currently disabled — not used for any account.
Privacy is built into the service, not added as a marketing claim.
View Trust & Privacy →FAQ
Yes — readable in your history, encrypted at rest. One toggle in Settings switches to hashed-only storage. Records auto-delete on your schedule — 365 days by default, adjustable down to 1.
Never. We verify the list, return the verdicts, and stop there — no selling, no sharing, no enrichment. Your data stays yours.
No. It's included in every plan, not sold as a premium add-on.
They never expire. No forced subscription, no use-it-or-lose-it.
Higher means more certain. Confidence is recalibrated nightly against real delivery outcomes — not a static model frozen at launch, and not a vendor marketing percentage.
Detected automatically. Set allow_async_followup on your request and we'll retry in the background once the delay clears — poll for the upgraded result instead of re-submitting the address yourself.
Both. Upload a CSV for bulk list cleaning, or connect Claude/Cursor via our MCP server — same API underneath.
Encrypted at rest. Deleted on your schedule. Your list is never our product.