How to read the result
Most verifiers answer with a single word and leave you guessing. This one returns the signals it acted on, so you can decide what an address is worth to you.
- Status — the actionable rollup, the one field to filter on:
valid,risky,invalidorunknown. - Score — a deterministic 0-100 confidence value derived from the signals in the same response. No model, no stored history, same input gives the same score.
- Result — the raw SMTP verdict behind the status:
deliverable,undeliverable,catch-allorunknown. - Reason — why the verdict is not plainly deliverable:
accept_all,no_mx,greylisting,provider_blocks_probe,invalid_syntax,timeout, and so on. - Did you mean — a suggested typo fix when the domain is one character away from a common one (
gmial.comtogmail.com). - Disposable, Free Provider, Role-Based — categorisation flags. A role address such as
contact@is often perfectly deliverable and still a bad idea to market to.
What the check actually does
Three layers, in order, stopping as soon as an address is provably bad:
- Syntax — catches malformed addresses before any network call.
- DNS — looks up the domain's MX records. No MX, no mail, whatever the mailbox.
- SMTP — opens a conversation with the mail server and asks whether the mailbox exists, then hangs up. No email is ever sent to the address you type, so nobody is notified that you checked them.
What it cannot tell you
Any verifier claiming 100 % accuracy is describing its marketing, not SMTP. Three cases are undecidable from the outside, and this one says so instead of guessing:
- Microsoft 365 and Outlook domains answer the same way for every mailbox, real or not. The reason comes back as
provider_blocks_probeand the status asrisky— that is the honest answer, not a failure. - Catch-all domains accept every address at the door and bounce later, so
catch-allmeans the domain is live but the mailbox is unconfirmed. - Greylisting asks any first-time sender to come back later. The result carries
greylistingand aretry_afterdelay.
Nothing you type is kept
The address is verified in memory and gone once the response is drawn. It is not logged, not cached, and not reused for a later request — including your own, a minute later. Every check runs on servers in France (OVH), so nothing leaves the European Union, and MailProbe is EU-operated rather than US-domiciled. Details on the privacy page.
Verifying more than one address
This page is capped at 15 checks a day per visitor, which is the right size for spot-checking a signup form or a suspicious address. Beyond that, a free account comes with 100 credits and opens two paths: a batch screen that takes a pasted list or a CSV file, and a REST API that verifies up to 500 addresses per request. Credits are pay-per-use, no subscription — see pricing.