Twenty-seven calling apps, one table

Every value in the table is taken from a structured research pass over each service's own published material, carried out in March 2026, and each classification is shown with the sentence it was read from on the comparison page for that service. Where the research found no published answer, the cell says so rather than guessing — “not published” is a finding, and filling it with a plausible value would make the whole table worthless.

This is deliberately narrower than a review. There is no scoring of call quality, no star rating and no opinion about which app is nicer to use, because none of those can be established from published material and all of them are what makes comparison articles unreliable. What can be established is what a service says it charges, what it says it requires, and where it says it runs — and those are exactly the things that decide a bill.

One field is our judgement rather than a quotation: whether a service can dial an ordinary telephone number. The dataset describes it in prose rather than recording it as a value, and it is the axis the whole table turns on, so it is stated per service with the description it was drawn from printed alongside on each comparison page.

Rates are not in the matrix, and that is a considered omission. Per-country rates change without notice, most of these services publish them only inside their own apps, and a table of prices captured on one date in one currency would be quietly wrong within weeks. The comparison pages quote the specific corridor rates that were published, with the date attached.

“Can this app call my mother's landline?” Half the apps people are told to use for international calling cannot dial a telephone number at all. That is not a defect — a messenger reaching another messenger for free is a better deal than any paid call, when both people have it. It is simply a different product, and comparisons that put messengers and calling services in one ranked list are comparing things that do not substitute for each other.

“What will this actually cost me?” The per-minute rate is the number every comparison publishes and it is not the number that decides the bill. A service billing per minute with rounding, plus a connection fee, can cost roughly double a service with the same headline rate on the short calls most people actually make. Those two fields are in this table and are in almost no other.

“What happens to money I do not spend?” Several well-known services void unused credit after a period of inactivity, and one of them published a 180-day rule. That is a real cost and it appears in no rate comparison anywhere, because it is not a rate.

We publish a calling app, so it is worth saying what that does and does not do to a table like this one. Telvio is the first row and is marked as ours. Its values come from the same rate sheet and product data the rest of this site is built from, and they are subject to the same rule as every other price here: nothing is typed by hand.

What we have not done is pick the fields to flatter it. The nine columns are the ones a bill actually depends on, and two of them — platforms and subscription — are ones where several competitors do better than we do. Telvio runs on fewer platforms than Skype, Viber, Telegram or Discord; it has no desktop client for Windows or Linux; it has no subscription for anybody who would prefer a fixed monthly cost; and it reaches fewer destinations than the large established carriers. All of that is in the table.

The comparison pages linked below go further, because they are about two services neither of which is ours. That is the honest test of a comparison: whether it is still useful when the publisher has nothing to sell in it.

Worth stating, because the fields that are missing are as informative as the ones that are here.

Call quality is not in the table and cannot be. The audio quality of an internet call is decided by the whole path — your connection, the route between the provider's carrier and the destination country, and the operator that finally delivers the call — and none of that generalises between destinations or between weeks. A rating of a service's call quality is a rating of somebody else's calls to somewhere else. The only test that answers the question you have is a call to the number you actually intend to call, which is why most of these services offer a free trial of some kind.

Support is not in the table. There is no way to establish from published material whether a company answers its email, and a claimed response time is not evidence of one.

Rates are not in the table, for the reason given above: most of these services publish prices only inside their own apps, and the ones that do publish change them without notice. Individual comparison pages quote the corridor figures that were published, with a date attached, and where a provider published nothing they say so — which is itself a comparison, and not a flattering one.

Nothing here is a recommendation. The table records what each service says it does. Which of them is right for you depends on who you call and how, and those are facts about you that no comparison can supply.