How do you check which chip you actually received?
Read a sample. A reading tool reports the chip family, the NFC Forum tag type and the free space within seconds, and comparing those three against what you ordered is the entire check. Do it on every delivery rather than once, because substitution happens between batches.
Read a sample
Take a tag from the delivery and read it with a tool that reports technical data rather than only content. The information is on the tag and takes no special equipment to retrieve.
Take several, from different parts of the delivery rather than from the top of one stack. A mixed shipment is a real possibility and a single sample cannot reveal it.
Do it before writing anything, since writing a batch you are about to reject wastes the effort.
What to compare
The chip, against the reference on the purchase order. This is the headline check.
The tag type, against what was specified. A change here is the most damaging possible substitution, because it can make the tags unreadable on phones regardless of everything else.
The free space, against the figure you sized your content on. A tag with the right name and less room than expected is still a problem.
And the identifier, which carries information about who manufactured the chip. A reading tool reports the manufacturer alongside the chip, and a mismatch between that and the claimed origin is worth investigating.
That mapping is registered under ISO/IEC 7816-6, whose first amendment established the IC manufacturer registration. It names the manufacturer, not the chip, so it narrows an investigation rather than settling it.
When it does not match
Stop before writing. Written tags are harder to return and a locked tag is not returnable at all.
Photograph the reading and the packaging together. A dated record of what arrived is what makes a conversation with a supplier short.
Decide whether the substitute is actually acceptable. A different reference with the same tag type and adequate free space may be fine, and that is your judgement rather than an automatic rejection.
And if it is acceptable, record it. A deployment where half the tags are one reference and half another is manageable when documented and confusing when not.
Making it routine
Check every delivery, not the first one. The first delivery is the one a supplier pays attention to.
Keep a record per batch: date, quantity, chip, tag type, free space, and a note of anything unusual. It takes a minute and it is what lets you answer questions two years later.
Keep one tag from each batch unwritten, labelled with the date. A reference sample is the fastest way to compare a suspect delivery against a known good one.
And read a sample again after any process you apply, since printing and lamination break tags that arrived perfectly good.
Frequently asked questions
Can I identify a chip without reading it?
Not reliably. Nothing visible on a finished tag identifies the chip inside, which is why a reading tool is the only practical check.
Does the identifier tell me the manufacturer?
Reading tools report a manufacturer alongside the chip, derived from the identifier. Treat it as a strong indication and confirm anything surprising with the supplier.
What if my tool reports a technology but no tag type?
That is the signal for a chip that is not a Forum tag type. Such tags may work in a closed system and will not read reliably on phones.
How many should I check?
A handful from different parts of the delivery. One tag cannot reveal a mixed shipment, which is the case worth catching.