Skip to the guide
Best Bag Authentication

The ranking / How we rank

How we rank bag authentication services

Our standards. Every figure is from the service’s own page, with the date we read it; quotation marks mean its exact words. One set of weights scores every service. Errors are corrected within two business days. How we rank.

The weights, out of 100
CriterionPoints
Price per bag27.5
Turnaround27.5
Certificate16.5
Verdict transparency16.5
Published brand count12
Total100

What a score is

A score is a sum. We read five things each service tells the public about itself, turn each into points with the rules below, and add the points up. The total is out of 100 and printed to one decimal. Each criterion’s points are rounded to two decimals first and the total is the sum of the rounded points, so the figures on an entry add up to its total exactly.

Every figure comes from the service’s own website, terms or app listing, and we print the page and the day we read it. Words between quotation marks are the service’s own; words without them are our record of its page. A figure a service does not publish is marked as not published and scores zero. We do not ask services for figures and do not estimate missing ones.

Who is ranked, and who is only listed

A service is ranked if it takes new customers and publishes a price for a single online check made from photographs. A service that does neither, or sells only to businesses, or comes bundled with a marketplace sale, is listed with its reason and gets no score. Listed and not ranked at present: Entrupy, Bababebi and eBay Authenticity Guarantee.

Price (27.5) and turnaround (27.5)

Price is what one bag check costs in dollars; turnaround is how long the service says that check takes, converted to minutes. The two are always read off one option the service sells, so a cheap tier’s price never sits beside a quick tier’s speed.

  • Every price and speed a service publishes for bags counts, across all of its price lists. Each pair is tried, the pair worth most points is kept (the cheaper one if two tie), and the entry names it.
  • A service that publishes prices and speeds without saying which price buys which speed is scored on its lowest price at the time it gives as standard.
  • A range of times is read at its slower end, so 24 to 48 hours counts as 48 hours. Where a service gives a standard time and a maximum, the standard time is scored and the maximum is shown.
  • Business hours and business days are read as ordinary calendar time, for every service.
  • Credits, tokens and multi-packs that any buyer can purchase are priced at their cheapest rate per check, rounded to the cent, and the entry says the dollar figure is our arithmetic. For example, LegitApp (rounded to the cent): it sells checks in tokens: this one takes 10 tokens, which come to $10.00 when bought in its 40-token bundle ($39.99).
  • Prices or speeds open only to members, to businesses or to volume buyers are left out for every service, and so are options a page lists but does not let a customer choose.

Points run along a straight line between the ends of the table. For price, the cheapest option any ranked service sells earns all 27.5 points and the dearest earns none; for turnaround, the fastest earns all 27.5 and the slowest none. Because the ends come from the options in the table, a service that adds or drops an option at either end moves every other service’s points too.

Zero therefore means one of two things. A figure a service does not publish scores zero. Separately, the dearest published price also scores zero, and so does the slowest published time, because each line runs from the cheapest or fastest option to the dearest or slowest.

Where the ends sit today: $10 (LegitApp) earns full price points and $120 (Authenticate First) earns none; 30 minutes (LegitGrails and LegitApp) earns full turnaround points and 2 days (Authenticate First) earns none.

Certificate (16.5)

  • A certificate issued with every check at no separate fee: 16.5 points. Where a service does not use the word free, the entry says included, no separate fee.
  • A certificate sold as a paid extra: 7.7 points, the same share of this criterion a paid certificate had under its first, 15-point version, rounded to one decimal.
  • No certificate, or nothing published about one: 0 points.

Verdict transparency (16.5)

We look for three things a service can tell you about how its verdict is reached, worth 5.5 points each:

  • that people examine the bag or decide the verdict;
  • what part AI or other technology plays, including a statement that none is used;
  • that you receive the reasons for the verdict as part of the standard check.

An element is credited only for the service’s own sentence, printed word for word on its entry with the page and the date we read it. Our summary of a service scores zero, and so does anything we do not print. Reasons given for one outcome only, such as a counterfeit result, count; reasons sold as a paid extra do not. Every service is searched the same way: its home page, questions page, process pages, price page and app listing. Each entry shows how many of these statements a service publishes and the sentences that earned them. Neither people nor software earns points by itself: what earns them is being told.

Each entry also has a short tag for who decides the verdict, drawn only from those credited sentences. It helps you read the entry and carries no points.

Published brand count (12)

The number of brands a service says it authenticates: its count for bags where it publishes one, otherwise its count across every category it covers, and each entry says which. Its own website is read before an app-store listing, and where it publishes several figures the lowest is taken. A list of names is not a count and scores nothing. The count is set against the highest count in the table (600 at present, LegitGrails), starting from zero, so half that count earns half the 12 points. It measures how many brands a service names, not how well it knows any one of them.

What we print but leave out of the score

  • Guarantees and refund terms, in full on each entry. Their conditions vary too much to turn into points.
  • What a service says about the people who check, which stopped earning points on 23 September 2026.
  • Anything a service says about using its paperwork in a marketplace or card dispute.
  • How you send a bag in, and any accuracy rate a service claims for itself.

One order on every page

The guide has exactly one order of services. Every page about a fashion house repeats it unchanged, because the house a bag comes from is not scored. Those pages add facts beside the order: the price each service publishes for that house by name, next to the price it is scored on, and whether its brand list includes the house. Leaving a house off a list costs a service no place.

The scale

Totals are out of 100 and shown to one decimal, so LegitGrails’ 95.38 appears as 95.4. Points per criterion keep two decimals. Services on an equal total are listed alphabetically; at present no two totals are equal.

Every score, worked out

The price and speed each service was scored on, then its points on each criterion. Add the points and you get the score; the score is printed with one decimal everywhere else in the guide.

Points by criterion for each ranked service.
Service Price /27.5Turnaround /27.5Certificate /16.5Verdict transparency /16.5Brand count /12 Score
1. LegitGrails 23.7526.6316.516.512 95.4
2. LegitApp 27.526.0516.516.56 92.6
3. CheckCheck 25.2523.1616.516.57.3 88.7
4. Legitmark 23.7526.0516.516.51.6 84.4
5. Real Authentication 1527.217.7113.4 64.3
6. Zeko's 17.513.8916.5110 58.9
7. Authenticate First 2013.897.7110 52.6
LegitGrails
$25 at 2 hours, from its Dior price list. Of the 32 price and speed pairs it publishes for bags, this one scores highest. 23.75 + 26.63 + 16.5 + 16.5 + 12 = 95.38
LegitApp
$10 at 3 hours, from its Handbag Authentication price list. It sells checks in tokens: this one takes 10 tokens, which come to $10.00 when bought in its 40-token bundle ($39.99). Of the six price and speed pairs it publishes for bags, this one scores highest. 27.5 + 26.05 + 16.5 + 16.5 + 6 = 92.55
CheckCheck
$19 at 8 hours, from its Handbags price list. It sells checks in credits: this one takes 20 credits, which come to $19.00 when bought in its 100-credit bundle ($95.00). Of the three price and speed pairs it publishes for bags, this one scores highest. Speeds sold only to paying members are shown and not scored: 4 hours, 24 credits (Premium users only). 25.25 + 23.16 + 16.5 + 16.5 + 7.3 = 88.71
Legitmark
$25 at 3 hours (listed as 3‑hour decision (typical)), from its Luxury & handbags price list. It is the one price and speed pair it publishes for bags. 23.75 + 26.05 + 16.5 + 16.5 + 1.6 = 84.4
Real Authentication
$60 at 1 hour, from its Basic Authentication price list. Of the six price and speed pairs it publishes for bags, this one scores highest. 15 + 27.21 + 7.7 + 11 + 3.4 = 64.31
Zeko's
$50 at 24 hours (listed as 24 business hours), from its Certificate price list. It sells this check in a pack of five for $250.00, which comes to $50.00 a check. It is the one price and speed pair it publishes for bags. 17.5 + 13.89 + 16.5 + 11 + 0 = 58.89
Authenticate First
$40 at 24 hours (listed as 24 hrs or less), from its Handbag authentication price list. That is the $20.00 Christian Dior check plus the $20 expedited fee for 24 hrs or less. Of the ten price and speed pairs it publishes for bags, this one scores highest. 20 + 13.89 + 7.7 + 11 + 0 = 52.59

The same ranking, with the prices services publish for one house by name: Hermès bag authentication · Louis Vuitton bag authentication · Dior bag authentication · Gucci bag authentication · Prada bag authentication · Saint Laurent bag authentication · Celine bag authentication.

Revisions

  1. We now print a service’s own sentence about the part technology plays in its checks, and credit it, whenever it meets the verdict-transparency rule: quoted word for word on that service’s entry, with the page and the date we read it. A re-read found that LegitGrails publishes such a sentence and that we had recorded it without printing it, so it had earned nothing. The rule credits an element when a service publishes it and this guide prints it, so leaving a sourced sentence out applied the rule differently to one service. On 16 November 2026 this change: it moved LegitGrails from second place to first in this ranking. It moved LegitApp from first place to second in this ranking.
  2. We read every service’s pages again and held every one of them to one written set of rules. A verdict element is credited only for a sentence the service publishes itself and this guide prints beside it, with the page and the date; our own summary of a service scores zero. Every price list a service publishes for bags now counts, rather than one list picked for it, and the highest-scoring price and speed pair is the one used. Membership, business and volume prices are left out for everyone, and credits, tokens or packs any buyer can purchase are priced at their cheapest rate per check. A range of times is read at its slower end, and a stated standard time is scored ahead of a stated maximum. A service is ranked only if it takes new customers and publishes a price for an online check from photographs; the rest are listed with the reason. We made the change because a review found the old method had been applied unevenly, in LegitGrails’ favour: its wording had been gathered from several pages while other services were read from one sentence or from our summary, its price came from a list chosen for it, and some figures for other services were out of date. On 1 October 2026 this change: it moved LegitApp from second place to first in this ranking. It moved LegitGrails from first place to second in this ranking. It moved Authenticate First from fourth place to seventh in this ranking. It brought Legitmark into this ranking, in fourth place. It took Bababebi out of this ranking (it had been in seventh place); it is now listed and not scored.
  3. A page about one fashion house used to keep its own order, re-scoring the services for that house alone. That let the guide answer one question two ways, so we stopped it: every brand page now carries the home page order and scores untouched, and the house’s prices and each service’s brand list sit beside that order as information, earning nothing. On 28 September 2026 this change: on the brand pages for Louis Vuitton, Gucci, Prada, Saint Laurent and Celine, it moved LegitGrails from second place to first.
  4. Reviewer depth, how many people a service says will look at a bag, stopped being a criterion, and the verdict element that asked for a second review went with it. A headcount says how a service staffs its checks, not what will happen to your bag, and services word it in different ways: some publish a minimum, some a maximum or the size of a team, some nothing, so the points were rewarding the wording. The ten points it carried were shared among the remaining criteria, which is how today’s weights came about: price per bag 27.5, turnaround 27.5, certificate 16.5, verdict transparency 16.5, published brand count 12. Verdict transparency now has three elements worth 5.5 each where it had four worth 3.75. What a service says about its reviewers is still printed on its entry. On 23 September 2026 this change: it moved LegitGrails from second place to first in this ranking.
  5. The first written version of these rules. From this date a price is only ever scored together with the speed it buys.