Sources & verification
Every fact comes from official sources and is stamped with a verification date. We re-check fees quarterly and document requirements twice a year. Each service page shows when its data was last updated.
Transparent: where the data comes from, how the score works, how we make money.
Every fact comes from official sources and is stamped with a verification date. We re-check fees quarterly and document requirements twice a year. Each service page shows when its data was last updated.
The 0–100 score is built from verified facts with published weights. No-DNI access and the real document list carry the most weight — our core focus. We never charge for ranking position.
The editorial score (from us) and the user score (from reviews) are calculated separately and never merged, so expert judgement and lived experience stay distinct.
Facts go stale, so each field has a re-check schedule: fees and rates quarterly, document requirements twice a year, features when a provider announces a change. Fields past their schedule are queued for re-verification and flagged on the page.
Anyone can report incorrect data. A report with a cited source becomes a correction proposal — it never overwrites live data. An editor checks the source before publishing, and every change is kept in a per-field history so you can see what changed and when.
Some links are referral links: if you open an account through them we may be paid. This never affects the score, the catalog order or the review text. Ratings don't depend on any partnership.
Each fact is tied to a source, ranked by authority:
A score is a claim, and a claim needs evidence. Every service climbs the same four steps; each one unlocks exactly one thing. Where the evidence is not there, we say so instead of publishing a number anyway.
A use case is a ranking lens, not a filter. The six dimensions stay the same; what changes is how much each one counts. Every profile adds up to 100%, so scores stay comparable across scenarios.
| Best for | Onboard | Reliab. | Fees | Support | Speed | App |
|---|---|---|---|---|---|---|
| Overall (site-wide) | 25 | 20 | 20 | 15 | 12 | 8 |
| opening without a DNI | 45 | 20 | 12 | 13 | 5 | 5 |
| receiving USD | 15 | 25 | 25 | 8 | 22 | 5 |
| holding USD | 12 | 40 | 20 | 10 | 8 | 10 |
| stablecoins (USDT/USDC) | 12 | 30 | 22 | 8 | 20 | 8 |
| crypto cashback | 10 | 25 | 30 | 8 | 10 | 17 |
| low fees | 10 | 20 | 45 | 8 | 10 | 7 |
| spending abroad | 8 | 20 | 28 | 14 | 12 | 18 |
| fast transfers | 5 | 25 | 15 | 8 | 40 | 7 |
| freelancers | 20 | 20 | 22 | 10 | 20 | 8 |
Figures are percentages. Bold columns are the dimensions a scenario leans on (25% or more): leaving one of them unrated caps the score for that scenario — missing data never counts in a service's favour.
A score is capped by its coverage: rated on a quarter of the weighted model, a service can score at most 25 out of 100, however good that quarter is. Filling a dimension can raise the ceiling; leaving one blank can never raise anything. Each score shows the number of dimensions it was rated on next to it.
A current account advertised at €0 a month, taken apart into the five charges that decide what it actually costs:
Same account, honestly stated: €0 a month, conditional on €700 sitting idle, plus €12 to have a card, plus €18 to leave. This is why we store the waiver conditions as part of the price, and why a fee a provider does not publish is recorded as not disclosed rather than as zero.
Country data answers three different questions — where a product exists, who may apply from here, and who actually gets in — and most published lists answer only the first. We store them separately. A service disappears from a country page only on evidence: an explicit exclusion by the provider, or absence from a list we know to be exhaustive. A list we merely hold cannot exclude, so a service we have not researched for your country stays visible, marked as unconfirmed, and ranked below everything we have confirmed. Being on a list is eligibility to apply; unless the provider states otherwise, it is never a guarantee of access.
A filter narrows a list to services that satisfy a condition; a ranking orders what is left by the published model. They are separate operations and we keep them separate: no filter reorders results, and no ranking silently drops a service that matched. A service missing the data behind a filter is excluded from that filtered list — an unknown value can never satisfy a filter — but it keeps its place everywhere else, and its profile says which fact is missing. A shortlist is a query result, not a recommendation.
323 of them official
Counted from the same data the pages render, at build time — not typed in by hand.
Fees, documents, IBANs, rails, deposit protection and the evidence states above are defined once in the glossary, with what each one means for you and how we treat it.
Affiliate disclosure: this site contains referral links. We may earn a commission, but it does not affect our ratings or recommendations.