VAT CSV syntax and duplicate checks before an official lookup.
For supplier finance and tax teams reviewing their own actual customers, the free public preflight tool checks format locally. It does not consult VIES, establish VAT validity, save a customer file or take payment.
- Open the workspace and choose a UTF-8 CSV, or paste it. Input stays in this browser during preflight.
- Select comma, semicolon or tab and enter the exact VAT-column header. Enter your own EU/XI requester number for format review only.
- Confirm that you may use the customer data, then choose “Inspect locally — free”. Review original cells, normalized identifiers, duplicates and rows needing format review.
- Use “Load synthetic example” to try fictional labels and documented integration-test numbers. No real registration claim follows from that example.
A reviewable handoff for your finance team
Keep the source export beside the preview. Confirm the exact VAT header and delimiter, compare each original cell with its normalized identifier, and investigate blank or malformed rows in the source system. A duplicate points to its first admitted row; it does not disappear. This public release keeps the review in this browser and offers no saved job or customer-file recovery.
When format review is finished, use an authorized official checking workflow under its purpose and data-use conditions. A local syntax pass does not establish registration, tax treatment or business legitimacy. Keep the official answer and its date separately from this local preparation.
Worked example: five rows, four syntax-admitted identifiers
In the workspace, choose “Load synthetic example”. It fills the comma-separated sample, VAT header and fictional requester number, then runs the local review. You should see five source rows, four unique admitted identifiers, one duplicate, no format-review rows and no blank VAT cells. No official lookup has run.
| Source row | Original VAT cell | Local interpretation |
|---|---|---|
| 2 | DE100 | Syntax admitted as DE100 |
| 3 | DE 100 | Normalized to DE100; same intended lookup as row 2 |
| 4 | DE200 | Syntax admitted; validity is unknown here |
| 5 | DE301 | Syntax admitted; no authority availability result here |
| 6 | DE600 | Syntax admitted; no authority availability result here |
“Admitted” describes this tool's basic syntax filter. It is not a country-specific checksum result or a registration answer. The separate synthetic outcome example gives DE200 a negative test answer and DE301/DE600 unresolved test answers. Those outcomes do not come from the local preflight and say nothing about actual customer registrations.
Two common CSV corrections
Header not found: a file beginning customer;vat needs the Semicolon setting and the exact header vat. A header written VAT requires VAT; case and spaces are preserved.
Unequal cell counts: in a comma file, a customer name containing a comma needs quotes, for example "Example, North",DE100. Keep the comma inside the quoted cell; do not remove customer text to make the row fit. If inspection fails, correct the source and inspect again. The old preview is cleared when input changes.
Semicolon exports and quoted customer text
This formatting illustration uses fictional labels and synthetic identifiers. It is not another customer list to attest to. Use the built-in synthetic button above for a ready-to-run walkthrough.
In a semicolon export, choose Semicolon and the exact header vat. A semicolon inside customer text belongs inside double quotes; a literal quote inside a quoted cell is written twice.
customer;vat
"Example; North";"DE 100"
"Example ""West""";DE200
The parser keeps two fields per row. The first customer cell becomes Example; North; the second becomes Example "West". Original VAT cells stay DE 100 and DE200; visible normalization admits DE100 and DE200 by syntax only. To correct your own export, preserve the customer text and repair the quoting or delimiter, then inspect again.
When input needs correction
Use country-prefixed EU/XI identifiers. GB and other registries are outside this tool. Ordinary spaces, dots and dashes are removed visibly; GR maps to EL. Repeated normalized identifiers share one intended lookup. Blank or malformed cells remain visible for review.
Limits: 256 KiB UTF-8 CSV, 1,000 data rows, 64 columns, 16,384 characters per cell, 250 unique admitted identifiers and 768 KiB expanded record. Keep a copy of your source file; this release does not save it for recovery.
Official checks and purchasing
Use the free official VIES service for actual checks. Read its purpose and data-use conditions. A format pass here is not an authority answer or a tax decision.
Private saved jobs, sign-in recovery and checkout are not open in the public preflight release. No payment is taken. The displayed pack price describes the proposed service; it is not an available purchase. Registered seller identity and public business contact are still being verified; no unverified mailbox or private owner address is published.