Bulk CSV Verification & Data Validation
One file, mapped for every kind.
Headers on row one. Every column is read for email, phone, WhatsApp and BIN values at once, so one spreadsheet can feed whichever check you need.
mapping is free · one file at a time · CSV or TSV
getting the file in
Drop it, read it, look before you spend.
Nothing is submitted by an upload. The file is parsed and previewed in the browser, and you decide what to verify from there.
the drop zone
Drag a CSV or TSV file here
or click to browse — headers on row one
Illustrative panel — the working uploader is in the dashboard. The button reads “Reading…” while the file is parsed.
what it accepts
- .csv and .tsv only — one file at a time.
- Headers on row one. The header row is what the hints are read from.
- 2 MB per file by default, and 1,000 rows per request.
- One cell can hold several values; they are split on commas, semicolons, pipes, tabs and newlines.
Mapping and preview cost nothing. Credits are charged per row, and only when you verify. The account still needs a positive credit balance to upload — that is a gate, not a charge.
detection
Automatic CSV column detection.
Detection reads the header row and the values under it to suggest which columns hold verifiable email, phone, WhatsApp or BIN values. It suggests; you choose what gets verified. It is not a guarantee — that is what verification is for.
Every column, scored against every check
A column qualifies on a header hint with at least one match, or on at least two matches at a 60% hit rate with no hint at all. Values that match nothing are left alone.
The rule, plainly
- Each column is scored against all three kinds at once.
- A column qualifies on a header hint with at least one match, or on at least two matches at a 60% hit rate.
- Header hints are read loosely — email and mail, phone and mobile and msisdn and contact, whatsapp and wa, bin and iin and card.
- The value test is a shape check: an address has an @ and a dotted domain, a phone number looks like a number, a BIN is 6–8 digits.
1,284 rows detected — 2 verification tools detected automatically.
One cell can hold several values
This is the feature people miss. A cell holding a list is not one address: the extractor splits on commas, semicolons, pipes, tabs and newlines, so a notes cell can still contribute. The preview shows every value found in a row, and each one you tick is verified as its own check.
what you can see go wrong
- Please drop a CSV or TSV file
- Only CSV and TSV files are supported.
- Could not parse the uploaded file.
- File exceeds the maximum limit of 1,000 rows.
- No verifiable email, phone, WhatsApp, or BIN values were detected in this file.
These are the uploader’s own messages, verbatim. The row limit in the fourth one is the deployment’s configured cap, not a fixed number.
What the answer looks like
The upload response carries the filename, the row count, the headers, a preview of the first rows, and — the useful part — which column was detected for each kind, how many rows matched, and which rows matched what. That is what the preview below is built from.
{
"filename": "leads-august.csv",
"totalRows": 1284,
"headers": ["email", "name", "notes"],
"detectedColumns": { "email": ["email"], "phone": [], "whatsapp": [], "bin": [] },
"suggestedColumns": { "email": "email" },
"matchedRowsByKind": { "email": 1102 }
}the preview
Pick the column. Tick the extras.
The selected column is tinted and its matches render in the accent. A row with matches outside the selected column gets checkboxes, so nothing is verified behind your back.
The preview
1,284 rows detected — 2 verification tools detected automatically. 1,102 email addresses selected for verification.select the email column
| name | notes | Row selection | |
|---|---|---|---|
| sarah@x.com | Sarah | — | Auto: sarah@x.com |
| — | Jordan | jordan@y.com | |
| — | Front desk | hello@z.com, info@z.com | |
| not-an-address | Sam | — | No Email match |
Illustrative rows. A value found in the selected column is picked first and reads Auto; values that live in another column stay unchecked until you tick them, and the column they came from is named beside each one. A row with nothing extractable reads “No Email match”.
the cost
A ceiling, not a bill.
You are quoted per selected row before anything runs. The number you see is the most it can cost.
estimated cost
The dashboard shows the maximum at the current rate for the selected rows.
- The estimate is a ceiling. A row that cannot be checked is skipped, and a request that fails is refunded.
- Phone, WhatsApp and BIN are billed at their own per-row rate, whatever quality you pick.
- Mapping, preview and column detection are outside the meter entirely.
verification quality
email onlyThe picker appears for email only. Choosing a higher quality changes the per-row cost, not the file you uploaded.
Known limits
A header row is required — the hints are read from it.
One file at a time. Upload the second one when the first is done.
Only extractable values count; a cell we cannot read is not a row.
A row with no matching value is skipped, not failed — and skipping is free.
Detection is a heuristic. A column whose values do not look like any supported shape is left alone, and you can point the run at the column yourself.
Mapping produces no verdict. Nothing here scores, ranks or validates a value — that is the verification step, and it is the only step that spends a credit.
Mapping is a read, not a verdict. Nothing on this page verifies anything — this step never spends a credit.
how it works
How CSV upload & mapping works.
One file goes in, and every column in it is read for the values it can hold.
- 01
One file, headers on row one
.csv or .tsv, one file at a time. The first row is read as headers — every column below it belongs to one of them.
- 02
Every column is scored for every kind
Each column is tested for email, phone, WhatsApp and BIN values using a header hint plus a value-shape check. A column qualifies on a hint with one match, or two matches at a 60% hit rate.
- 03
A cell can hold more than one value
Values inside one cell are split on commas, semicolons, pipes, tabs and newlines, so a notes column can still contribute. The values you tick are the values that get verified.
- 04
You choose what gets verified
A preview of the first rows shows the selected column tinted, with matches outside it as per-row checkboxes. A value in the selected column is picked first; the others wait for you to tick them.
- 05
Nothing is charged yet
Mapping and preview cost nothing. The quote before a run is a maximum for the rows you selected, not a charge.
Mapping, drawn
Needs a header row, one file at a time. Only extractable values count, and a row with no matching value is skipped rather than failed.
What is bulk CSV verification?
Bulk CSV verification takes a spreadsheet you already have and checks the values inside it, instead of asking someone to type addresses, numbers or BINs in one at a time. You upload a CSV or TSV, and MultiValidator reads the columns for the supported value types.
Detection is a suggestion, not a verdict: you see which column was detected for each kind, preview the extracted values, and choose what enters the run. Nothing in the file is verified — and nothing is charged — until you select rows and start verification.
How bulk CSV verification works.
Six steps from a file on your disk to per-row verdicts. The first five are free; the sixth is where credits are spent.
Upload your CSV or TSV
One file, up to 2 MB and 1,000 rows by default, with the header row on line one. An account needs a positive credit balance to upload — a gate, not a charge.
Detect the supported values
Every column is scored for email, phone, WhatsApp and BIN values by header hint plus value shape. Detection suggests which columns hold verifiable values.
Preview and review
The response carries the headers, a preview of the first rows, the detected columns per kind, and how many rows matched each kind.
Select what to verify
Pick the column, then tick the per-row extras. A detected value is never verified behind your back — you choose what enters the run.
Review the estimated cost
Before a run you see a ceiling: selected rows × unit cost, multiplied by the email quality you chose. Skipped rows and failed requests are not charged.
Run verification
Selected rows go to the pipeline for their kind and come back with a verdict, a reason and — for email — a score. Results land in your history for download.
Step six is documented per kind: bulk email runs queued and is polled to completion, while bulk phone and bulk BIN return completed in the same response.
What can you verify from a CSV?
Three value types, detected from the header row and the shape of the values. A column is only considered if its values actually look like one of them.
Email addresses
The full pipeline: syntax, DNS/MX, disposable, the SMTP mailbox probe, spam-trap, role and free-provider flags, then a verdict and a score.
How email checks work →Phone numbers
Parsed to E.164 against the numbering plan, then resolved for line type, carrier and risk signals from provider data.
How phone checks work →BIN / IIN values
Six to eight digits, resolved to the card brand, type and issuing bank where the provider data covers the range.
How BIN checks work →Detection reads both halves: a header that says email or bin or mobile, and values that match a supported shape. A cell can hold a list — split on commas, semicolons, pipes, tabs and newlines — and each value you select is verified as its own row.
CSV upload requirements.
What the uploader accepts today, with the limits as configured in this deployment.
- File type
- .csv or .tsv, matched on the file name
- Files per upload
- One — the uploader returns when the job finishes
- Header row
- Required on row one; it is what the hints are read from
- File size
- 2 MB by default (BULK_MAX_MB)
- Rows per file
- 1,000 by default (BULK_MAX_ROWS)
- Account state
- A positive credit balance to upload — a gate, not a charge
The two deployment settings are environment values, not per-account entitlements — self-hosted installs set their own. Headers are required because every hint is read from that first row.
How bulk verification credits work.
Reading a file is free; checking a row is not. The meter starts at the verification step and never before it.
What each step costs
- Upload
- Free — screened for type, size and rows
- Detection & mapping
- Free — nothing enters the meter
- Preview & selection
- Free — review and ticking cost nothing
- Verification
- Charged per checked row at the current rate
- Email quality
- Multiplies the per-row cost for email
Standard, high and extreme email checks are billed at the current per-row rates set by the administrator.
The quote is a ceiling
Before a run you see the maximum it can cost: the rows you selected, multiplied by the unit cost for that kind. Phone, WhatsApp and BIN rows are billed at their own per-row rate. Nothing is deducted at that point — the estimate is a promise about the upper bound, not a charge.
- A row that cannot be checked is skipped, and a skipped row is never charged.
- A request that fails is refunded, so a provider failure is not billed to you.
- The estimated maximum is shown before the run, next to the rows it applies to.
CSV mapping is not verification.
This is the distinction the whole page rests on. Everything before a run is reading; the verdict comes from the checks themselves.
- Mapping identifies values that may be worth checking — it does not check them. A detected address is a candidate, nothing more.
- Detection is a heuristic on headers and shapes. It can be wrong in both directions: a column it leaves alone, a column it picks that you would not.
- Preview shows you the extracted values, so the decision about what enters the run is yours rather than the detector's.
- Verification is the only step that produces a verdict, and the only step that spends a credit. Until you start a run, no value has been checked.
- A detected email is not a valid email, a detected phone number is not a reachable one, and a detected BIN is not a live card — those are verification outcomes, with a reason attached.
When you are ready to check the values, the outcomes are documented per kind: email validation, phone validation and the IIN/BIN checker each explain what a verdict means and where it stops.
CSV verification vs. manual lookup.
The same checks either way. The difference is what you have to do to run them over a list.
manual lookup
- One value per lookup, typed by hand
- Repetitive for anything list-sized
- Hard to review a large dataset in one pass
- Selection is manual and easy to skip
- Total work is hard to estimate up front
bulk CSV verification
- Upload the spreadsheet you already have
- Supported columns detected for you
- Extracted values previewed before a run
- Rows selected and ticked before anything runs
- A maximum credit cost quoted up front
- Results returned per row, with the reason attached
Both paths run the same pipeline, so a row checked in a file is verified exactly as a row typed by hand would be. If you would rather drive it from a script, the API documentation covers the upload, job and download endpoints for lists.
Bulk CSV verification use cases.
The jobs a list-shaped dataset is usually handed to MultiValidator for. Every one of them ends in a verdict and a reason, not a guarantee.
Email list cleaning
Check addresses before a send or an import, so the dead rows are found while they are still cheap to drop.
CRM data hygiene
Review an imported contact file for supported value types before it lands in the CRM.
Lead list validation
Screen a lead export for email and phone values, with the reason code on every row that fails.
Phone data validation
Resolve the phone numbers already in your spreadsheets for line type, carrier and risk signals.
BIN data processing
Read BIN and IIN values out of a structured dataset and resolve them against the provider data.
Bulk verification API
Drive the same files from your own pipeline: upload, poll the job, download the results as CSV.
Read the API docs →Looking for the single-value versions? Email validation, phone validation and the IIN/BIN checker run the same checks one value at a time.
Bulk CSV verification FAQ.
What the uploader accepts, what mapping costs, and what it does not do for you.
CSV and TSV. The upload route accepts a file named .csv or .tsv and rejects anything else, one file per upload.
Stop mailing ghosts.
Start with free credits. No card, no call, no “book a demo”.