MEFERI ME87: the handheld with RFID inside the processor

What the ME87 is, and what it is not

One point of spelling first, because search queries and price lists get it wrong: the manufacturer is called MEFERI, not MIFERI. MEFERI Technologies Co., Ltd. is based in Chengdu in Sichuan province, not, as is often assumed, in Shenzhen. Alongside the ME87 the company sells the ME61, ME63 and ME74 handhelds, the MP30 mobile printer and the MH50 headset. Anyone searching for “MIFERI ME87” means this device.

The ME87 belongs to a class of device that did not exist two years ago: a handheld that looks from the outside like a slightly thick smartphone but carries enterprise engineering inside — a replaceable battery with hot swap, IP68, a drop rating of 1.5 metres, a real scan engine instead of camera decoding, and 5G with Wi-Fi 7 on top. Every figure on this page is a manufacturer statement. We have measured no device, and our research turned up no independent measurements to hold the manufacturer's figures against. That is not an accusation; it is a risk factor you have to price in when comparing.

We now have two manufacturer sources rather than one: the official specification sheet held by this project, and MEFERI's public ME87 page, which we re-read on 01.09.2026 for this English edition. Where the two disagree we follow the newer one and say so in the text. Where the public page repeats what the datasheet says, that is a repetition by the same party and not confirmation by anyone else.

One distinction matters for placing the device: the ME87 is at heart a mobile computer with an RFID function, not an RFID reader with a computer function. That sounds academic and decides everything about how it is used. A Zebra RFD40 or a TC22R is built around the RFID antenna: a large antenna area, a pistol grip, a trigger, everything laid out for continuous capture. The ME87 is built around a display and a mobile radio, and the RFID function makes use of what is there anyway. That is elegant, and it has physical consequences we take apart in the next section.

Who this device would suit, if you could buy it

The honest answer is a shop that mostly needs near-field work: goods-in against the delivery note, marking up a new collection, hunting for one particular piece, weekly spot counts of a single rail. Add the payment angle, and a boutique with one till could plausibly run price marking, counting, searching and card acceptance on a single 290 gram device. That is a genuinely attractive picture, and we set it out at length under practice rather than dismissing it.

The device would suit a business that counts stock continuously in small bites instead of once a year in one long day. If your stocktaking model is a full count with the shop shut, the range works against you, and the arithmetic for that is below as well.

What the ME87 explicitly is not

It is not a fast full-stocktake device. At two metres nominal, and one to one and a half metres among textiles on our own estimate, you have to bring the reader to every piece of stock. A device in the six-metre class sweeps a whole clothes rail from the aisle. That single difference is worth around an hour and a half in a shop of 120 square metres, and no firmware update or training changes it.

It is not a procurable product in Europe today. That is the sentence a buyer needs, and it stands here rather than in a footnote. There is no price, no distributor and no documented European radio approval, so no quotation can be calculated, no delivery date promised and no service case planned. The detail sits under availability and approval.

It is not stock control software either. It delivers EPC numbers, and what becomes of them is decided by the application behind it. And it prints nothing: new season goods are tagged beforehand on an RFID printer that writes text, barcode and chip in one pass. How the label is married to the existing barcode is set out in our guide to RFID stocktaking; the basics are in the RFID guide and on the page about RFID readers.

RFID inside the processor: what that changes

To see why that is a break with how these devices have been built, it helps to look at the two established arrangements first.

Arrangement one: the clip-on sled

The classic route. A mobile computer you already own has a module clipped on or paired over Bluetooth that holds the entire RFID part: the reader chip, the power amplifier, a large antenna in the grip, its own battery. The Zebra RFD40 is the textbook example, and the Chainway R6 is the budget version of the same idea. The advantage is performance: a dedicated, large antenna produces range. The disadvantage is the two-device logic — two batteries, two cradles, a matching adapter, two firmware states — and around 780 grams in the hand.

Arrangement two: the integrated module

The second route builds the RFID reader into the device permanently, usually a module of the Impinj E710 class with its own antenna in the housing. The Zebra TC22R is built that way, and so are the Chainway C72 and the Chainway C5. One device, one battery, one cradle — and still high performance, because the antenna is given room by the design. The price of that is bulk, weight and power draw. A module of that kind draws current even when it is merely on standby.

Arrangement three: the engine inside the processor

This is where the ME87 starts. MEFERI describes the Qualcomm QCM6690 as the first processor architecture with fully integrated UHF RAIN processing: no external decoding module, with the signal processing running in blocks that already sit on the chip itself. That claim of being first is a manufacturer claim. We could not verify it independently and reproduce it as such.

The manufacturer's public page goes further and calls the ME87 the world's first handheld with a built-in UHF RFID reader. That broader claim cannot be read literally. Handhelds with a built-in UHF reader have been on sale for years; the TC22R and the C72 on the pages linked above are two of them. Read charitably, the claim means what the datasheet says more precisely: not the first handheld with an integrated reader, but the first with the reader inside the application processor rather than in a module beside it. We report the claim because it is on the manufacturer's page, and we do not adopt it, because the plain reading is not true.

What follows from the architecture is easy to follow. First, power: what the main processor handles in passing needs no supply rail of its own and no standby current. MEFERI names exactly that as the benefit, without giving figures. Second, space: a module that does not exist needs no board. That is why the ME87 weighs 290 grams with its battery, where a TC22R is 569 grams and a handheld plus sled comes to roughly 780 grams. Third, latency: there is no interface between a reader chip and an application processor across which data must be pushed. Whether that shows up in practice, the datasheet does not say, and we cannot put a number on it.

The antenna is the catch

The real compromise is in the datasheet too, just quietly: reading happens over the shared WWAN antenna, the antenna the device needs for 5G and LTE. That is exactly where the figure of two metres comes from. An antenna designed primarily for mobile network bands does not reach, in the UHF RFID band, the gain of one built for the purpose, and a housing 16 millimetres deep has no more room to give. The datasheet does not supply that explanation — it is our own reading of the physics, not a manufacturer quotation.

Against that MEFERI sets an extension. Over the ten-pin POGO connector at the back, the datasheet says, a compact UHF antenna can be attached that turns the device into a full RFID terminal. That is the most interesting sentence in the whole document and the least supported. What range that antenna reaches, the datasheet does not say. No product name, no part number and no price appear either, and it does not show up in the list of optional accessories — single-slot cradle, four-slot cradle, four-slot battery charger. That antenna is therefore not demonstrably available. While that remains the case, plan with two metres.

A footnote on the processor name, because it causes confusion. For the ME87 MEFERI names the Qualcomm QCM6690 (FP1), octa-core at 2.0 GHz on a 4 nanometre platform. For the forthcoming Zebra TC201, Zebra names the Qualcomm Dragonwing Q-6690. The designations plainly sit close together, and Qualcomm introduced the Q-6690 platform only in October 2025. Whether the two are the same silicon cannot be established beyond doubt from the public documents, so we do not treat the names as equivalent. Note also that the manufacturer's public page names no processor model at all, only a “new Qualcomm processor”; the model number on this page comes from the datasheet alone.

Range, tags per second, frequencies

Two metres does not sound dramatically short at first. In practice, though, it is an open-field figure under favourable conditions: one label, a clear line of sight, no attenuation. In a fashion shop you work against textiles that carry moisture, against chromed metal rails that reflect, and against densely stacked goods where the outer labels shadow the inner ones. Two metres of open field then becomes one to one and a half metres realistically, and that figure is our own estimate rather than a manufacturer statement. This is no fault of the device; it is physics, and it applies to every UHF system. It simply weighs more heavily here, because there is less to start with.

The read rate of 200 tags per second, by contrast, is a solid figure, and it is one of the few the manufacturer states in both sources. For comparison: the Zebra EM45 RFID manages 20 tags per second, the announced Zebra TC201 around 200, and the TC22R and RFD40 up to 1,300. The ME87 sits solidly in the middle — and, this being the point of the page, that is entirely enough. Read rate is almost never the bottleneck in a shop stocktake. We do the arithmetic under practice.

On anti-collision, since the question comes up: EPC Gen2 is a time-slot procedure. Labels do not answer simultaneously but one after another, in windows the reader sets. What a datasheet gives as “tags per second” is therefore always a throughput under ideal conditions with a known tag population. Where very many labels sit very close together, real throughput falls, because more collisions have to be resolved. At 200 tags per second the headroom is large enough that this plays no part in fashion retail.

Every figure in this table comes from the manufacturer: from the official MEFERI ME87 Specification Sheet held by this project, or from MEFERI's public ME87 page retrieved 01.09.2026. There is no independent verification of any of them. Where a source is silent, the table says so in the status column rather than leaving the row out.

Why “860 to 960 MHz, region dependent” is not a specification

That phrase describes the whole worldwide UHF range, and the worldwide range is not one band but several mutually incompatible ones. Europe uses 865 to 868 MHz. North America uses 902 to 928 MHz. Japan, Australia and China each use something else again. A device stated as 860 to 960 MHz is not thereby stated as legal in Europe; it is stated as belonging to a family of builds of which some are and some are not. Which build you would receive is precisely what neither source says, and it is the first thing to settle in writing. The manufacturer's public page is worse on this point than the datasheet, not better: it gives no frequency figure at all.

Full technical specification

Two rows changed when we re-read MEFERI's public page on 01.09.2026 for this English edition. The operating system moved from Android 16 with an upgrade path to Android 18, which is what the datasheet says, to Android 15 with an upgrade path to Android 20, which is what the manufacturer now publishes. The scan engine moved from a firm SE4780 to no public commitment at all. Both readings stay visible above, because a buyer holding the older datasheet needs to know the two documents disagree.

What actually counts in this specification

Three points lift the ME87 above the usual level, independently of the RFID question.

First, the drop test. Nearly every manufacturer states a drop height. The addition “across the whole operating temperature range” is rare — it means the 1.5 metres also hold at −20 °C, where plastics become brittle. That MEFERI cites a certified test report for it is a good sign. We do not, however, hold that report.

Second, hot swap with a 30-second buffer. A built-in 90 mAh cell keeps the device alive while the battery is changed. That sounds like a detail and matters in practice: you swap the battery mid-count without the session dropping, without a restart and without logging in again. Together with over 80 per cent of capacity after 500 cycles, that means a realistic service life of several years before replacement batteries fall due.

Third, the software runway — and this is where the sources moved. The manufacturer now publishes Android 15 out of the box, an upgrade path as far as Android 20 and up to five years of Google security patches. Android 20 is several years out, so the far end of that promise cannot be verified today by anyone, and we present it as a manufacturer commitment rather than a property of the device. Even so, the near end of it is unusually generous. For comparison: on the TC22R, Android 16 is the last supported version, and even that figure comes from distributors rather than from Zebra's official version table.

Two things this table does not contain, because no source states them: a transmit power in dBm or ERP, and any statement about whether the device can write tags. On a device sold into fashion retail, the second matters more than it looks. Writing is what lets you correct a mis-written label at the counter. We flag it as not found rather than assuming a capability that is standard elsewhere.

In a fashion shop: 4,000 pieces over 120 square metres

Let us work it through properly. Assume a typical owner-run fashion shop: 120 square metres of sales floor, around 4,000 pieces in stock, of which roughly 3,200 are on the floor and 800 in the stockroom, in reserve under the tables and in the window. On the floor stand around 60 goods carriers: round rails, double rails, wall rails with several levels, centre tables with stacked jumpers, plus a few shelf walls.

Step one: the pure read time

4,000 labels divided by 200 tags per second is 20 seconds. Even if you generously assume half the throughput, because labels are captured several times over and collisions have to be resolved, you land under a minute. For comparison, a device managing 1,300 tags per second would need three seconds. The difference between 20 seconds and 3 seconds, in a stocktake that runs for hours, is of no consequence whatever. Remember that number: anyone comparing handhelds on read rate is comparing the wrong quantity.

Step two: the walking, at two metres

At two metres nominal — in practice more like one to one and a half among textiles — you can no longer capture a round rail from the aisle. You have to walk round it once, take the device past at hip height and again at head height, and fan the goods briefly where a rail is densely packed. A top shelf at two metres you reach only at arm's length. A stack of jumpers on a centre table you sweep from two sides, because the outer pieces shadow the inner ones.

Put 60 to 90 seconds per goods carrier against that, including moving position and glancing at the counter. At 60 carriers and 75 seconds on average, that is 75 minutes of sweeping the sales floor alone. Add the stockroom and reserve with around 800 pieces, usually packed more densely and therefore slower, at an estimated 30 to 40 minutes. Window, fitting rooms and the till area add another 10. And the follow-up work — hunting for pieces that were not read, checking labels, resolving discrepancies — realistically 30 to 60 minutes.

Total: around two and a half to three and a half hours for one person. That is a projection of ours on the basis of manufacturer figures, not a measurement.

Step three: the same shop at six metres

With a device in the six-metre class, such as the RFD40, the movement changes fundamentally. You walk the aisle and swing the device; a round rail is captured in passing, and so are the upper shelves, because you no longer have to get to them. Instead of 75 seconds per carrier, 20 to 30 will do, and many carriers are picked up incidentally. 75 minutes becomes around 25. With the stockroom, the side areas and the follow-up work you land at one to two hours — the figure we work with at HANGCOUNT, and also a projection of ours rather than a guaranteed value.

Read times are calculated from the manufacturers' tags-per-second figures. Every other line in this table is our own estimate for a shop of this size and layout, not a measurement and not a manufacturer statement. A shop with wider aisles and fewer, larger carriers moves faster; a densely packed one with high shelves moves slower.

A second cross-check: a working figure for a handheld with adequate range is an effective capture of around 700 pieces a minute in a live shop, and that is a manufacturer figure. 4,000 pieces would on that basis be through in just under six minutes of actual capture. That figure points the same way: the electronics are never the problem. The person doing the walking is the time budget.

What two metres is entirely enough for

This is the other side, and it deserves the same space. There is a whole set of jobs in a shop where a short range is not a drawback but an advantage, because you want to capture one particular piece and not accidentally half the next rail.

• Goods-in: a carton is opened and its contents read against the delivery note. The carton is in front of you, and two metres is more than enough.
• Marking up a new collection: attach the label, marry it to the existing barcode, move to the next piece. Pure near-field work.
• Finding an article: a customer asks for a size 38 in blue. You start the search and walk the floor until the tone gets louder.
• Spot counts: one carrier, one product group, one shelf — once a week instead of once a year. This is where a light device with a long battery genuinely beats a pistol grip.
• Taking payment: with EMVCo-capable NFC the same device could in theory accept the card as well. More on that next.

The till argument: one device for stock and for payment

One point deserves attention of its own, because it is economically interesting for small shops with a single till. The ME87's NFC module is an NXP PN7221, and the datasheet sets it out expressly as EMVCo 3.x contactless compliant, with support for Apple ECP 1.0 and 2.0 including VAS. The manufacturer's public page goes on to describe the device as SoftPOS-ready in those words. That is the hardware basis for card payment directly on the device, with no separate terminal.

Think that through for a boutique: one device with which you mark up new goods in the morning, count stock at midday, find a piece from the stockroom for a customer in the afternoon and take her card at the fitting room. That saves not only a terminal but a walk to the till.

Two qualifications belong with it. First, hardware is only the precondition; whether a particular payment application may be operated on a particular device is decided by payment service providers and certification bodies. No such approval for the ME87 is known to us, and none is claimed in either source. Second, the caveat from the next section applies here too: a device you cannot obtain in Europe solves no till problem.

Availability, approval and what we could not find

This section deliberately sits ahead of the price and comparison sections, because it qualifies everything else. A device with excellent figures that cannot be delivered, and whose radio conformity for Europe is not evidenced, is simply not usable for a purchase today — however well it reads on paper.

The rows marked “not found” record the result of a search, not a statement by the manufacturer. We looked for each of them expressly on MEFERI's public ME87 page on 01.09.2026, and we report the absence rather than smoothing it into a softer form. An absence is not proof that something does not exist; it is the state of the public record on the date given.

Why the missing radio test standard is the most important point

The European standards named in the datasheet are all safety and EMC standards. EN 60950-1 is the old equipment safety standard for information technology. EN 55032 and EN 55024 govern emissions and immunity, EN 61000-3-2 and -3-3 the effects on the mains supply. None of those says anything about whether a transmitter in the UHF RFID band meets European requirements. ETSI EN 302 208 is the standard that does, and it is not in the list.

That is not proof that the device is non-conformant. Datasheets are incomplete, and a manufacturer that states CE has to meet the Radio Equipment Directive. It is, though, exactly the gap you cannot leave open on a UHF device — all the more so when the same document gives the band as “860 to 960 MHz, region dependent”, covering the entire worldwide UHF range including the American 902 to 928 MHz. A device in a US configuration may not be operated in the European Union at all.

And the newer source does not close the gap; it widens it. MEFERI's public ME87 page names no CE marking, no Radio Equipment Directive, no harmonised standard and no frequency. On a page whose whole purpose is to persuade integrators and distributors to register for evaluation units, that is a conspicuous silence. We record it as a fact about the public record on 01.09.2026, without speculating about the reason.

What the early-access programme actually offers

To be fair to the manufacturer, the page does not pretend to be a shop. It says plainly what it is. Registration gets you documentation, a place in the pre-order queue, direct technical support during testing, and what the page calls a special discounted evaluation offer for a test sample. Samples ship worldwide by DHL or another express service, and the manufacturer states that registration does not guarantee a sample, because availability is limited.

Read commercially, that describes a manufacturer at the start of a launch, selling evaluation units directly out of Chengdu to integrators and distributors who might later resell. That is a perfectly normal stage in a product's life. It is simply not a stage at which a single fashion shop should be buying, because there is no price to budget, no supplier to hold to a delivery date, and no European service address should the device fail in the second year.

The second thing that counsels caution is the timeline. Qualcomm introduced the processor platform in October 2025. Devices built on a platform barely a year old are typically still near the start of their approval and distribution chain. That FCC authorisations are documented for the older MEFERI models ME61 and ME63 under grantee code 2A9LJ, while none can be found for the ME87, fits that picture: the device is real, but it is early.

What this device is not good for

The most honest sentence about this device is this: it fails not on its engineering but on its range and its availability. Neither can be compensated for by software, by training or by a cleverer way of working.

Not suitable for the annual full stocktake of a sales floor. Two metres forces you into a way of working in which you walk round every goods carrier. It works, but on our projection it takes one and a half to two times as long as a device that reaches six metres. In a shop that closes for the count, that is money.

Not suitable for high shelving and deep stacks. Anything you cannot reach by hand — upper shelves, stockroom racking, packed reserve — becomes climbing or becomes incomplete at two metres. This is exactly where greater range earns its keep.

Not suitable as a calculable purchase. Without a price, without a distributor and without a documented EU radio approval, no quotation can be calculated, no delivery date promised and no service case planned. If you need a solution for the next stocktake, you cannot wait for this device.

Only partly suitable for hours of continuous capture. A flat device with no grip and no trigger is held differently from a pistol shape. 290 grams is light, but the way you hold it while sweeping rails is more tiring over hours than the weight suggests. Whether MEFERI offers a grip attachment is not apparent from either source.

And one caveat that runs under all of the above: every figure here is a manufacturer figure. With established manufacturers you can check a datasheet against distributor statements, test reports and regulatory databases. With the ME87 you currently cannot, because outside the manufacturer's own documents there are almost no dependable sources. Where we have found two sources, they are both MEFERI, so they corroborate each other in the way a person corroborates themselves. That is not an accusation, and it is a risk factor to price in.

None of that makes the engineering less interesting, and this page would be dishonest if it pretended otherwise. RFID inside the application processor is the first structural change in this device class in years. If it becomes normal, every enterprise handheld becomes an RFID device, and the entire section you are reading gets rewritten. That is a statement about the next few years, not about your next stocktake.

Price and the costs around it

That is unusual, and it is the only honest answer available. An estimated figure would live on in price comparisons and quotations with nothing underneath it. What we can supply instead is the range that comparable devices occupy, so that you can judge whether waiting for this one is worth it.

All euro figures are net list prices excluding VAT, with the source and date given in each row. EU business buyers with a valid VAT identification number are invoiced under the reverse-charge procedure, so the net figure is what you budget with. Project prices frequently sit below list. US prices are quoted for orientation only and are expressly not converted into euro: they contain neither European taxes nor import duty, freight or a dealer's margin. Availability and configuration are to be checked with the dealer before any order.

What falls due alongside the device

The handheld is rarely the largest item in an RFID rollout. Anyone comparing device prices alone underestimates the project regularly. These blocks come on top:

• Labels. Around 12 to 15 cents a piece in small quantities in Germany (Karley, Zebra UCODE9, as at 08/2026). For an initial stock of 4,000 pieces that is 480 to 600 euro net for the legacy stock alone, then a running cost for every new item. More on our page about RFID labels and tags.
• An RFID printer. You cannot do without a device that prints a label and encodes it in the same pass. What happens technically and what matters is on our page about RFID printers.
• Charging. On the ME87 the single-slot cradle, the four-slot cradle and the four-slot battery charger are optional, so not in the box. No prices are published for any of them.
• A spare battery. On a device with hot swap at least a second battery makes sense, or you are throwing the feature away.
• Software and the interface. A handheld without stock control software counts numbers, not articles. The handover to your existing till system usually runs over CSV or Excel.
• Set-up and handover. A device profile, Wi-Fi, a label layout, a test run. The handover itself takes around half an hour in our experience — tagging the legacy stock at 200 to 300 pieces per person per hour is the real work.
• On the ME87 additionally: import, customs, a conformity assessment and a service concept for a device with no European distribution. None of those can be quantified today, and that is itself the answer to whether the device is buyable.

One thing worth saying plainly about what RFID buys you, whichever device you end up with. Stock accuracy in fashion retail without RFID sits at 65 to 75 per cent and with RFID at 93 to 99 per cent, with a sales effect of up to +5.5 per cent (Beck, 2018, ECR Shrinkage & On-Shelf Availability Group / GS1 UK). Those are the figures that carry the business case. They do not depend on which of the devices on this site you buy — which is precisely why buying an unobtainable one makes no sense.

Accessories, charging and software

Charging and battery logistics

Charging runs over USB-C 3.0 with Quick Charge 4.0 at up to 27 watts, or over the two-pin POGO connector at 5 volts and 3 amps — the latter being the route the cradles use. A full charge takes up to two and a half hours per the manufacturer. In the box, per the datasheet, are an 18 W power supply with four plug variants, the 5,000 mAh battery, a USB-C to USB-C cable, a hand strap and a screen protector.

For a single shop with one device the single-slot cradle is enough. As soon as two people count in parallel, or several devices are in use, the four-slot version earns its place — not for the charging time but because a fixed place to put things down is what actually gets devices onto power at night. The four-slot battery charger charges spare batteries separately from the device and, combined with hot swap, is the clean answer for stocktaking days: one battery in the device, one charged beside it, a change without the session dropping.

The external UHF antenna

The ten-pin POGO connector at the back is the most interesting expansion option on this device. Per the datasheet a compact UHF antenna can be attached there that makes the ME87 a full RFID terminal. If that antenna lifted the range well beyond two metres, the whole assessment on this page would be different. That is exactly what the document is silent about: no range figure, no product name, no part number, no price, and no appearance in the list of optional accessories. The manufacturer's public page does not mention it at all. While that remains the case, plan with two metres.

The manufacturer's software

MEFERI names three software building blocks: the OmniSoftware Suite, the OmniIndustry App Suite and Shadowalk as an MDM or EMM solution for device management. Shadowalk is the one the manufacturer's public page puts its name to, alongside Android AER. On what any of them actually does we hold no dependable information. Whether a publicly documented RFID SDK exists, through which stock control software could address the read engine, is not apparent from the datasheet. For integration with an existing merchandise system that is the decisive question, and it would have to be settled before any purchase.

For comparison: Zebra publicly documents a complete toolkit with 123RFID Desktop, its RFID SDKs and the connection to SOTI Connect. In a software integration that is a concrete advantage which appears on no datasheet as a technical figure. The manufacturer's page does say the device supports modern Android APIs and SDKs for NFC, barcode scanning and MDM or EMM — note that RFID is not in that list.

On service, MEFERI names twelve months of hardware warranty from the date of shipment and the extended MeCare programme. How a service case would actually run in Europe — the address to send to, the turnaround, whether a replacement device is provided — is unresolved without a European distribution partner. Twelve months is short for this class of device in any case: the TC22R comes with two years of manufacturer warranty at its European distributor.

What to buy instead, by name

Range figures are manufacturer statements and open-field values. On the TC22R the European spec sheet says 24.4 m and the US document for the same device says 12.2 m; we work with the conservative figure. The Chainway figures are catalogue values measured with large anti-metal tags, and a third to a quarter of them is realistic in a shop — that reduction is our own estimate, not a manufacturer figure. In a shop, between textiles and metal, the practically relevant range of every device here is well below its catalogue number. An overview of all of them is in the RFID hardware section.

Buy this instead

If you want to be able to count stock this year, buy the Zebra TC22R. Around 1,295 euro net at a European distributor, one device, one battery, one cradle, no adapter matching, and a documented European band of 865 to 868 MHz at 1.45 to 31.45 dBm ERP in the manufacturer's own spec sheet. Two years of manufacturer warranty. It is the boring choice and it is the right one.

If you already have handhelds in the shop, look at the Zebra RFD40 as an attachment. From 839 euro net for the standard variant, roughly six metres of range, 1,300 tags per second. Have the distributor confirm compatibility with your specific device model, because for several Zebra models it is not documented.

If the budget is tighter, look at the Chainway C72 at around 942 euro net or the Chainway C5 at around 992 euro net. Both have a European dealer and a documented European price, and both do the full-stocktake job the ME87 cannot. Read the caveat on their own pages first: Chainway names ETSI EN 302 208 nowhere either, so ask the supplier for the EU declaration of conformity. The difference from the ME87 is not that the paperwork is perfect — it is that there is a European supplier to ask, a price to compare, and a device you can have next week.

If you have a phone you want to keep using, look at the Chainway R6 at around 659 euro net — the cheapest route into a full stocktake in this section.

If what you really need is single-article capture and a phone, the Zebra EM45 RFID in a smartphone form factor and the forthcoming Zebra TC201 are the ME87's nearest relatives — similarly short range, but with a route to market. Note that the TC201's RFID variant is announced for Europe only at the end of 2026, so it shares part of the ME87's problem.

If you want to read RFID at the till, there is a device of its own for that in the Zebra DS9908R, whose range of roughly 46 centimetres is kept deliberately short so that the next table is not charged to the customer in front of you.

And the ME87? It is a device to watch, not a device to buy. Put it on the list for the generation after next. If MEFERI, or anyone else, ships a processor-integrated reader with a European band configuration, a declaration of conformity naming the radio standard and a European distributor, this page will be rewritten and the verdict with it. Until then the interesting architecture and the unbuyable product are the same object, and both statements have to be made in the same breath.

Radio approval: what applies in Europe

The European frame

UHF RFID readers in Europe operate in the band from 865 to 868 MHz. That band is harmonised across the European Union and the European Economic Area through CEPT recommendation ERC/REC 70-03 on short range devices, and the technical basis is the harmonised standard ETSI EN 302 208 in its version V3.4.1 of December 2023, which permits up to 2 watt ERP for interrogators in this band. A further band from 915 to 921 MHz with up to 4 watt ERP is available in addition; which parts of it are released nationally differs between member states.

Because the band is harmonised, the same device is lawful in every EU and EEA country. What differs is the national regulator that administers the general authorisation: the Bundesnetzagentur in Germany, ARCEP in France, UKE in Poland, RDI in the Netherlands, and so on for the other member states. The list is longer than that and there is no point reproducing it here; the point is that the frame is common and only the administering authority changes.

For normal operation within that general authorisation, no individual licence and no registration is needed, and no spectrum fee is payable. You buy the device, you switch it on, and you keep the conformity papers. Responsibility for operating within the frame sits with the operator, that is with your business, not with the dealer.

The United Kingdom sits outside this. Since leaving the EU it applies its own rules, with Ofcom Interface Requirement IR 2030 for licence-exempt short range devices and UKCA marking in place of CE; if you are buying for a UK site, check the current position with your supplier rather than relying on the European frame described here.

A frequent misunderstanding concerns the comparison with the United States, where 4 watt EIRP applies. Because ERP and EIRP differ by 2.15 dB, 2 watt ERP corresponds to roughly 3.3 watt EIRP. The American limit is therefore only about 20 per cent above the European one, not double, as is occasionally written. That conversion is our own. Anyone claiming European devices have half the range of American ones is not converting the units carefully.

What MEFERI documents, and what that means for a commercial buyer

This is the sore point of the page, and it needs stating in the terms a business buyer works in rather than in the language of a specification. The datasheet lists CE among its certifications, but for Europe it names safety and EMC standards only: EN 60950-1, IEC 60950-1, EN 55032 Class B, EN 55024, EN 61000-3-2 and EN 61000-3-3. ETSI EN 302 208 is not among them. The UHF band is given as “860 to 960 MHz, region dependent” — the full worldwide range, with no European configuration set out. No transmit power in dBm or ERP is given either. And the manufacturer's public page, which is the newer document, names no CE marking, no Radio Equipment Directive, no harmonised standard and no frequency at all.

Here is what that means if you buy this device for a business. You, as the operator, are responsible for operating within the general authorisation. The document that lets you demonstrate you are is the EU declaration of conformity, which names the standards the manufacturer applied. For a radio device, the standard that matters is the radio one. If the declaration you are handed does not name EN 302 208, you have a CE mark and no evidence about the transmitter — and the transmitter is the whole point of the purchase. In a dispute, in an insurance question, in a tender, or if a regulator ever asks, the CE logo on the housing is not the answer; the declaration is. This is the same caveat this site puts on the Chainway pages, and here it is sharper: Chainway at least declares Directive 2014/53/EU in its manuals and a European distributor holds CE and RoHS certificates you can download. For the ME87 there is neither a declaration to read nor a European supplier to ask for one.

None of that is evidence that the device is non-conformant. It is evidence that conformity for Europe is not currently demonstrable from the public record, and for a commercial buyer those two things have the same practical consequence: you cannot sign for it. A private individual might take the risk. A business that has to document what it operates cannot.

What CE means and what it does not

The CE marking is the manufacturer's declaration, or that of the responsible economic operator, that the device meets the applicable EU rules, in the case of radio devices above all the Radio Equipment Directive 2014/53/EU. It is not an official test and not a quality seal. For you as the operator it is nonetheless the decisive basis, because without it the device may not be placed on the EU market at all. The accompanying paper is the EU declaration of conformity, you are entitled to ask for it, and it is the document in which a reference to EN 302 208 would appear.

Something else that is no evidence of EU conformity: an FCC authorisation. That is a United States authorisation and says nothing about European conformity. In this case the point is moot in any event, because no FCC authorisation for the ME87 could be found at all — only for the older ME61 and ME63 under grantee code 2A9LJ.

The awkward part: builds for other regions

A device configured for 902 to 928 MHz may not be operated in the European Union, not even briefly for testing. That band is allocated differently here, among other things to mobile network services. With most manufacturers you can read the region from the part number: Zebra uses -E8 or -WR for Europe and worldwide and -US or -FT for the United States and Canada, and Chainway dealers label the European C72 as C72-V4/CM710 UHF ETSI. For the ME87 no such region coding is published at all, which is a further reason to have the band configuration confirmed in the order documents rather than inferred from a part number. Be particularly alert with marketplace listings and direct imports, where stock built for other regions is regularly sold with no note.

The basics of frequencies, power limits and the difference between ERP and EIRP are set out at greater length in the overview article RFID explained, and what part the antenna plays is covered under RFID antennas.

This section summarises the position as at 01.09.2026 and is no substitute for legal advice. Particular installations, such as fixed doorway antennas at high transmit power, carry additional requirements, and national rules on siting can apply.

Frequently asked questions about the MEFERI ME87

Read on

Sources and the date of the figures

Every figure about the MEFERI ME87 on this page is a manufacturer statement, from one of two MEFERI documents, and neither has been independently verified. Where we have calculated, estimated or placed something, the text says so. Where a figure is missing, uncertain or corrected against the German version of this page, the text says that too. The manufacturer's public page was retrieved on 01.09.2026; the specification sheet and the comparison sources on 26.08.2026, with the Chainway prices from 31.08.2026.

• MEFERI, public product page “ME87 Introduction”: meferi.com, retrieved 01.09.2026. The newer of the two manufacturer sources and the one this page follows where the two disagree. The source of Android 15 with AER and GMS, of the upgrade path stated as far as Android 20, of up to five years of Google security patches, of the confirmation of 200 tags per second, of roughly 2 m of range, of IP68 and the 1.5 m drop, of 5G and Wi-Fi 7, of the NXP PN7221 with MIFARE, FeliCa, Apple ECP and EMVCo and the description of the device as SoftPOS-ready, of the open wording on the scan engines with an optional SE4770 for long range, of the “special discounted evaluation offer” for early-access members, of worldwide sample shipping by DHL or another express service, and of the statement that registration does not guarantee a sample. Checked on the same page and not found: any price, any European distributor or sales office, any CE marking, any reference to the Radio Equipment Directive or to ETSI EN 302 208, and any frequency figure at all, including the European band of 865–868 MHz.
• MEFERI ME87 Specification Sheet (English), the official manufacturer document held by this project, retrieved 26.08.2026. The source of every detailed figure on this page: the integration of the UHF engine into the Qualcomm QCM6690, ISO/IEC 18000-63 and EPC Class-1 Gen2 V2, the band statement of 860–960 MHz region dependent, the maximum range of 2 m over the shared WWAN antenna, the display, the memory, the camera, the battery with its 90 mAh buffer cell, the dimensions and weights, IP68, the drop, tumble, vibration, thermal shock and ESD figures, the temperature and humidity ranges, the connectors, the audio and sensor lists, the security features, the listed certifications and European standards, and the 12 months of warranty with the MeCare programme. It also carries the Android 16 and Android 18 figures that the manufacturer’s public page has since superseded.
• Qualcomm, “Dragonwing Q-6690 Processor Product Brief”: docs.qualcomm.com (PDF), retrieved 26.08.2026. The source of the platform introduction date of October 2025. This page does not treat the Q-6690 and the QCM6690 as the same silicon, because the public documents do not establish that.
• ETSI EN 302 208 V3.4.1 (2023-12), radio equipment for identification in the bands 865–868 MHz with up to 2 W ERP and 915–921 MHz with up to 4 W ERP: etsi.org (PDF), retrieved 26.08.2026. Cited on this page as the legal frame for operating a device in Europe, and expressly not as a manufacturer statement about the ME87.
• CEPT recommendation ERC/REC 70-03 on the use of short range devices, annex on non-specific short range devices and RFID — the instrument through which the 865–868 MHz band is harmonised across the EU and the EEA. Retrieved 26.08.2026.
• National general authorisations administered by the member states, for example Bundesnetzagentur in Germany, ARCEP in France, UKE in Poland and RDI in the Netherlands. Retrieved 26.08.2026. The basis for the statement that no individual licence and no registration is needed for normal operation.
• Ofcom, Interface Requirement IR 2030 — the United Kingdom regime for licence-exempt short range devices, outside the EU framework since the UK left the Union, with UKCA marking in place of CE. Retrieved 26.08.2026.
• GS1, “Regulatory status for using RFID in the UHF spectrum”: gs1.org (PDF), retrieved 26.08.2026. The worldwide overview of frequency allocations behind the statement that 860 to 960 MHz spans several mutually incompatible regional bands.
• Zebra Technologies, spec sheet “TC2xR Series”: zebra.com (GB), retrieved 26.08.2026. The source of the comparison figures for the TC22R: up to 1,300 tags per second, the European band of 865–868 MHz at 1.45 to 31.45 dBm ERP, and a weight of 569 g.
• Zebra Technologies, spec sheet “RFD40 UHF RFID Sled”: zebra.com (GB), retrieved 26.08.2026. The source of the comparison figures of roughly 6 m and 1,300 tags per second.
• Price sources for the comparison tables, all net: MicroNordic for the TC22R at around 1,295 euro, ico.de for the RFD40 from 839 euro and for the Premium and Premium Plus variants, AutoID Warehouse for the TC201 at 648.83 to 881.34 euro, all as at 08/2026; kaufid.de for the Chainway R6 at 659.10 euro and the Chainway C5 at 991.50 euro and shopid.eu for the Chainway C72 at 941.60 euro, both retrieved 31.08.2026 and both shops operated by ITFutuRe s.r.o., Czech Republic; Atlas RFID Store for the EM45 RFID at 1,138 USD as at 08/2026, quoted for orientation only and expressly not converted into euro. Details and the limits of each figure are on the individual device pages.
• Label prices in small quantities in Germany, around 0.12 to 0.15 euro a piece: Karley, Zebra UCODE9, as at 08/2026.
• Stock accuracy without RFID 65 to 75 per cent, with RFID 93 to 99 per cent, sales effect up to +5.5 per cent: Beck (2018), ECR Shrinkage & On-Shelf Availability Group / GS1 UK.
• Our own projections on this page, marked as ours wherever they appear and never presented as manufacturer figures: a stocktake of two and a half to three and a half hours at 2 m and one to two hours at 6 m for 4,000 pieces over 120 square metres; a realistic range of 1 to 1.5 metres among textiles; 60 to 90 seconds per goods carrier at 2 m against 20 to 30 seconds at 6 m; tagging legacy stock at 200 to 300 pieces per person per hour; and a handover of around half an hour. None of these are measurements.

All figures are as at 01.09.2026. Prices and availability change; check both with the dealer before ordering. HANGCOUNT has no sales relationship with the manufacturers or dealers named and does not sell the devices described at the prices given.