Loading...
Every six hours we record what each money transfer provider would actually deliver to a recipient, on every corridor we track. The archive behind these studies holds 3,290,393 quote observations going back to March 10, 2026. These 7 datasets are what comes out of it — each with its method, its limits, and the figures it actually supports.
No provider is cheapest everywhere. Wise is the most frequent winner on 44 of 212 corridors, and on 25% of corridors the provider that is cheapest today is not the one that usually wins.
125,025 provider-day observations across 212 corridors with a genuine comparison, over 91 days. A day counts only when two or more providers quoted.
Banks cost 1.40× what specialist apps cost on the same corridors — 3.23% of the amount sent against 2.31%. Against the cheapest provider on the same corridor, a bank transfer loses 2.61% on average.
208 bank quotes and 363 specialist quotes at $1,000, across 98 corridors and 31 banks.
Specialist providers cost an average of $30.23 in fees and exchange-rate markup on a $1,000 transfer; banks cost $55.30.
61 providers priced across 1,084 corridors carrying at least one quote at the headline amount — a wider base than the 212 corridors with a sustained multi-provider record. True total cost = the gap between what a mid-market transfer would deliver and what the provider actually delivers, so it captures the fee and the markup together.
Weekends are marginally cheaper, not dearer: a mean markup of 1.51% at weekends against 1.59% on weekdays, a difference of 0.073 percentage points. Which provider you choose is a far larger lever than which day you send.
3,290,393 quote observations from 328 snapshots across 84 providers, 2026-03-10 to 2026-09-07, corrected for uneven sampling with outliers quarantined rather than dropped silently.
Sending $100 costs 3.75% of the amount against 2.55% at $1,000, and the cheapest provider is not the same at both — InstaReM at the smaller amount, LemFi at the larger.
17 providers with 20+ usable quotes at both amounts. No $10,000 tier is published — only 11 providers quote at that amount, roughly one per corridor.
Of the 203 corridors where enough providers compete to score one, 64% currently read "good" or better as a time to send.
A 0–100 rating from four weighted components, computed from a median of 164 days of recorded provider rates per corridor. Single-provider corridors are excluded from the distribution rather than counted as neutral.
What a transfer is worth on arrival depends more on local price levels than on the fee. The study pairs our live transfer costs with World Bank PPP conversion factors across 83 countries.
World Bank PPP and GNI indicators, refreshed against live quotes. The interactive version is the salary-abroad calculator.
These datasets are published under CC BY-NC-SA 4.0. Quote the figures, reproduce the tables, and cite SendMoneyCompare with a link to the study page. Each study states the date of the data behind it — please cite that date rather than the date you read it, because the figures move with the scrape.
Raw data. 3 of these studies publish their underlying table as a CORS-open CSV, so a figure can be checked against the row that produced it: Provider Consistency Index, Bank vs App Transfer Cost Index, Transfer cost by amount. There is also a public quotes endpoint for AI assistants.
Method and corrections. How the underlying quotes are collected and costs calculated is set out in our methodology. Where we have got something wrong and fixed it, it is logged in corrections. For questions about a dataset, or a cut of it we do not publish, get in touch.
Analysis by
Akif HazarviEditor-in-Chief
Akif Hazarvi is the editor-in-chief of SendMoneyCompare with 8+ years in fintech and cross-border payments.
Corrections and independence: figures move with the scrape, so cite the data date above rather than the date you read this. Where we have published something wrong and fixed it, it is logged in corrections. No provider can buy a ranking or a place in these datasets — see our editorial policy and methodology.