Bulk Email Verification
Every row, every stage.
The whole pipeline runs per row, and every stage hands back the reason it produced. This is the queued job: submit it, watch it advance, then download the verdicts.
queued · every stage · a reason on every row
what runs per row
The stages, per row.
The same eight stages a single check runs, applied to every row in the file — with the stage outcome and its reason kept on the row.
Syntax
RFC-5321/5322 parsing, normalization, length limits. Zero network calls.
DNS / MX resolution
A live lookup for MX and A records. No records, no mailbox: mx_not_found.
Disposable detection
Known throwaway providers matched at the domain level before any probe.
SMTP mailbox probe
A polite MAIL FROM / RCPT TO exchange — nothing is ever sent. Hosts may refuse; when they do, you get the raw response code.
Spam-trap heuristics
Flags for trap-pattern addresses before they poison your sender reputation.
Role-account detection
info@, sales@, support@ — addresses that rotate, get shared, or go unread.
Free-provider flag
gmail.com, outlook.com and friends bounce at higher rates; flagged as context, not a verdict.
Verdict + score
deliverable / risky / undeliverable / unknown, a 0–100 score from the stage outcomes, and a didYouMean typo suggestion when one applies.
Email runs every stage. Phone, WhatsApp and BIN run shorter pipelines — they do not reach the SMTP stage.
how the job runs
Queued, polled, downloaded.
Email is the job that queues. You submit, a job comes back, progress is polled until it completes, and the results download as a file.
while it runs
The counts the dashboard shows mid-job, and the pills it ends on.
verdicts and score
Four verdicts, one score.
For email the score is a real composition: the stage outcomes allocate a hundred points, and the marker shows where the row landed.
four verdicts
where those scores come from
The score starts as the sum of the seven stage weights below, and is then held inside the band its verdict allows: 80–100 deliverable, 40–70 risky, 0–15 undeliverable, 0–40 unknown. A greylisted probe sums to 80, which is why it lands at 70 — the top of the risky band. It is a quality score computed from the checks that ran, not a prediction about a send.
Every other stage passed and the probe was greylisted rather than refused, so the row is risky and the score sits at the band's ceiling.
What a bad list does to you
Bulk email runs thousands of rows at once, so the stakes scale with the file. Sending to addresses nobody cleared runs down your sender reputation rather than one bad inbox.
quality
Quality, and what it costs.
Standard, high or extreme — email only, and the only thing it changes is the per-row cost.
Higher quality buys a deeper provider check on each row, not a different pipeline. Phone, WhatsApp and BIN are billed at their own per-row rate whatever you pick here.
Known limits
The SMTP probe only fires once syntax and DNS pass — a domain with no mail route never reaches it.
A throttled, greylisted or timed-out probe does not come back as a clean deliverable: the row keeps the provider's status, so it lands risky or unknown rather than being forced into a verdict.
A catch-all server accepts a probe for an address that may not exist. The server is flagged; its mailboxes are not readable from the outside.
Re-running a row costs another credit and can legitimately land differently: greylists clear, providers throttle, a domain's records change.
The score is a quality score computed from the stages that ran, held inside the band its verdict allows — not a prediction about a send.
Charged per row at the current rate, only when a row is checked. A skipped row is never charged and a failed request is refunded.
What is bulk email verification?
Bulk email verification checks a list of addresses as one job instead of one address at a time. Every row runs the same pipeline a single check runs, and every row comes back with its own verdict, its own reason and a score — so a list can be read row by row rather than trusted as a whole.
On MultiValidator the email job is the one that queues. You submit the list, a job is created, the job is polled while it advances, and the verified results download when it completes. Nothing about the run is a claim that mail will arrive — it is a record of the checks that ran and what each one found.
how it works
How bulk email verification works.
Every address in the file runs the full pipeline, and the job runs on its own until the verdicts are ready to download.
- 01
Provide the addresses
A CSV or TSV file with a column mapped to email, or a JSON array of addresses through the API. Both create the same kind of job against the same pipeline.
- 02
Choose the quality once
Standard, high or extreme applies to the whole run for email and sets the per-row cost. There is no quality picker on phone, WhatsApp or BIN.
- 03
Submit the list
The job is created as pending with the rows it will check. The row cap is the deployment's BULK_MAX_ROWS setting — 1,000 rows by default — and credits are spent per checked row, never at submission.
- 04
Poll while it advances
Progress is read from the job rather than guessed: processed rows and the four verdict counts, updated as the run moves.
- 05
Review the results
Deliverable, risky, undeliverable or unknown per row, each with the reason the stage produced and a 0–100 score held inside the band its verdict allows.
- 06
Download the list
When the job reports completed, the verified CSV downloads from the dashboard. The job status endpoint returns the same per-address results to API clients.
Bulk email, drawn
The SMTP probe fires only after syntax and DNS pass, and a throttled, greylisted or timed-out probe keeps the provider's status rather than being reported as a clean deliverable. Re-running a row costs another credit and can legitimately land differently.
Working from a spreadsheet? The file path — headers, column detection and the preview — is documented on CSV upload & mapping. Phone, WhatsApp and BIN jobs return completed in the same response instead of queuing.
What each check proves — and what it does not.
Eight stages run on every row. Each one answers a narrow question, and the verdict is only as strong as that answer.
- localSyntax
- That the mailbox exists. Syntax only says the address is well-formed after normalization; nothing has been queried yet.
- live DNSDNS / MX resolution
- That the domain accepts mail for a particular address. MX records describe where mail would be accepted, not that it is.
- provider listDisposable detection
- That a domain outside the throwaway list is safe to send to. Detection matches a known-provider list at the domain level, and the list is not exhaustive.
- SMTP handshakeSMTP mailbox probe
- That the mailbox will keep accepting mail. The probe reads the host's answer at that moment, and nothing is ever sent.
- heuristicSpam-trap heuristics
- Anything about the person behind the address. Trap-shaped patterns are a reason to leave a row out, not evidence about intent.
- heuristicRole-account detection
- That a role address is dead. info@ and sales@ usually exist — they simply tend to be shared, rotate, or go unread.
- provider listFree-provider flag
- That a consumer-provider address is worse. The free-provider flag is context for the list, not a verdict on the row.
- computedVerdict + score
- The future of the mailbox. The verdict and score summarise the stages that ran, against the answers given at that moment.
What email verification cannot guarantee.
A verdict describes a check, not a mailbox forever. These are the things the pipeline does not claim, on this page or anywhere else.
- That a message sent later is delivered. The checks ran at a point in time, against the answers the providers gave then.
- Inbox placement. Filtering happens at the receiving end, after the verification is already finished.
- Engagement. Whether a person reads, opens or replies is not a property of the address.
- Long-term availability. Mailboxes are abandoned and retired after a check has run.
- Human ownership. The pipeline checks an address; it does not establish who is behind it.
The same caveats are documented for the single address check, because both paths run the same pipeline.
Bulk email verification use cases.
The jobs a list is usually handed to MultiValidator for. Each one ends in a per-row verdict and reason, not an outcome promise.
Email list cleaning
Check a list before a campaign or an import, so the rows that fail are found while dropping them is still cheap.
CRM data hygiene
Verify the addresses in an imported contact file before they land in the CRM, with a reason attached to every row.
Lead list checking
Screen an export for deliverable addresses and read the reason code on the rows that failed.
Marketing database upkeep
Re-check the segments that matter before a send, and see the maximum credit cost of the run first.
Spreadsheet jobs
Already working in CSV? Map a column and submit the file instead of building an array by hand.
CSV upload & mapping →Bulk email API
Submit an array of addresses, poll the job, and take the per-address results back into your own pipeline.
Read the API docs →Bulk email verification vs. a single check.
Neither path is better than the other — they run the same pipeline. What differs is how much you hand over at once.
single email verification
One address per request, verified in the same response. It fits inside a form submit, a signup flow or a webhook handler, where waiting on a job would be the wrong shape.
The eight stages and the reasons are identical — the email validation page documents what comes back.
bulk email verification
A list per job, submitted as a file or an array, processed as one queued run and polled to completion. Every row keeps its own verdict, reason and score, and the whole list downloads as CSV.
That is the shape for cleaning an existing list, where the row count is the point and per-row reasons are what the review needs.
Bulk email verification FAQ.
What the job accepts, what the verdicts and the score mean, and what none of it promises.
Bulk email verification checks many addresses as one job instead of one at a time. Every row runs the same pipeline a single check runs and comes back with its own verdict, its own reason and a score.
Stop mailing ghosts.
Start with free credits. No card, no call, no “book a demo”.