Which country to pick for SMS verification: when the flag decides it, and when the route does

The short version

For the large majority of the 133 services in the catalogue, the country of your number is a price decision and nothing more. The platform wants to prove a human holds a reachable line; it does not care whether that line sits in Lagos or Lisbon. Where the country does matter, it matters completely — a service that is not offered in a market can accept the code and then refuse the account minutes later, and no route tier repairs that. So the first question is never "which country is cheapest" but "does this platform care where the number is from at all".

The second thing to hold onto is that the price spread across countries measures demand and carrier capacity, not quality. A $0.09 average across all 133 services in Nepal and a $5.90 average in Singapore describe two different wholesale markets, not a cheap line and a good line. Forty of the 116 countries here average $0.20 or less across the entire catalogue, fifty-four land between $0.20 and $1, and twenty-two sit above $1. Nothing in those bands predicts whether a given platform will accept the range.

And the third: when a code fails, the route tier is almost always the cheaper thing to change. The catalogue carries 400 live routes — 118 virtual, 161 physical and 121 premium — and the tier is what a platform's carrier lookup actually reads. Moving from virtual to physical in the same country costs a multiplier of 1.20× to 1.80× on a base price that is often measured in cents. Moving country costs you the entire attempt and tells you nothing, because you changed two variables at once. The delivery guide works through the failures themselves; this one is about the map they happen on.

These look like seven different problems and are really three: the platform's own coverage, the route under the number, and a price you read at the wrong level. The first row is the expensive one, because it is the only failure here that a refund does not cover — the code arrived and was billed, and the account still did not open.

The thread running through all seven is that an instant refusal is about the range, a long silence is about the route, and a failure that happens after the code lands is about the country. Only the last of those is a reason to change the map. The other two are answered one tier up, in the same place, for a few cents — and if nothing arrives at all, the 20-minute window returns the money on its own.

Why one code is $0.21 in one country and $47.77 in another

The catalogue holds 15,428 live prices — every one of the 133 services in every one of the 116 countries — and they run from $0.01 to $47.77, averaging $0.69. That is not one product with a wild markup. A price here is the cost of a carrier route in that market multiplied by how hard registrations from that range are fought over, and both halves move by an order of magnitude. Nepal and Myanmar average $0.09 across all 133 services because mobile capacity is genuinely cheap there. Singapore averages $5.90 because it is a small, expensive market whose ranges are in permanent demand.

The spread also depends far more on the service than the country. WhatsApp is the most expensive service on the whole map, averaging $3.19 and reaching $47.77 in Singapore; Telegram averages $1.99 and tops out at $19.91 in the same country; most of the catalogue never leaves single-digit cents. Of the 15,428 pairs, 6,579 cost $0.20 or less and only 2,595 cost more than $1. If your service is one of the quiet ones, the country you pick barely changes your bill — and if it is WhatsApp, the country is the single biggest number in it.

What none of this measures is whether the range will be accepted. A dear country is dear because other people want its numbers, which is a demand signal and not a quality one — and the ranges people want most are also the ranges platforms scrutinise hardest. There is no price at which a virtual route stops resolving as cloud-issued. The country directory lists every market with its live routes and multipliers, and the live feed publishes actual delivery and stock movement rather than a claimed success rate.

Price is the column everyone reads and the least informative one on this table. Read the tier column instead: it tells you how many moves you have if the first code fails. Nepal in the highlighted row is one of the two cheapest markets on the whole map and still carries all three tiers on three routes — which is why the cheapest country is very often a perfectly good answer.

Two patterns are worth carrying away. Countries with five routes — Canada, France, Germany, Indonesia, Italy, the Netherlands, Poland, Russia, Spain, the United Kingdom and the United States — give you the most room to escalate, and they are also the ones platforms watch most closely. And a low price does not imply a thin map: 98 of the 116 countries carry all three tiers, including most of the cheapest ones. Every country page shows its own routes with the multiplier beside each.

The tier moves more than the flag

Every route in the catalogue is one of three kinds, and the kind is what a platform's carrier lookup returns before any SMS is queued. A virtual route is cloud-issued and resolves as such; a physical route terminates on real SIM hardware and comes back as an ordinary mobile subscriber; a premium route draws on the pools held back for the checks that refuse everything else. The split across the 400 live routes is 118 virtual, 161 physical, 121 premium, and the price of the difference is a multiplier: 0.85× to 1.00×, 1.20× to 1.80×, and 1.80× to 2.50× respectively.

Put those two scales next to each other and the arithmetic is decisive. A premium route in a country where the base price is $0.12 costs about $0.30 and resolves as cleanly as anything on the site. A virtual route in Switzerland, where the catalogue averages $3.92, costs many times that and still resolves as cloud-issued. Escalating the tier is almost always cheaper than escalating the country, and unlike a country change it tests exactly one thing — which is the only way to learn anything from a failure.

The exceptions are worth knowing before you rely on the pattern. Five countries carry no virtual route at all — Japan, Singapore, Thailand, Turkey and Vietnam — so their entry price is already a physical route. Nine carry no premium tier, including India, Indonesia, Russia and Turkey, so physical is the top of the ladder there. And seven carry no physical route, which means the escalation goes straight from virtual to premium with nothing in between.

Coverage is not availability, and neither one is acceptance

Three separate things get compressed into the phrase "does this country work", and separating them saves most of the wasted attempts. Coverage is whether we can sell you a line there — all 116 countries carry all 133 services, so on this site coverage is never the constraint. Availability is whether the platform itself operates in that market. Acceptance is whether it will take that particular range. A country can pass the first two and fail the third, and a country can pass all three today and fail tomorrow, because ranges are recycled and reputations move.

Availability is the one that bites hardest, because it fails late and silently. A regional bank, a national delivery app, a marketplace that only ships domestically — these will happily send a code to a number in a market they do not serve, and then refuse the account at the next step, or months later when they re-check. The code is billed because the code arrived. Nothing here refunds a verification that worked exactly as promised into a platform that was never going to keep the account.

There is also a quieter version of the same problem. Registration country often sets an account's region, language, currency or store, and for some platforms that is decided once and hard to move afterwards. If you are verifying something you intend to keep, spend thirty seconds checking whether the app lets you change region later. If it does not, choose the country you actually want to live in rather than the one that is nine cents cheaper — and check the service page for how that platform behaves before you buy.

Renting changes which country you want

A one-time activation exists for twenty minutes, which makes the country a throwaway decision: it has to clear one check and then it is gone. A rental runs 7, 14, 30 or 90 days from $4.20, and for that whole term the line is yours alone and nobody else can be sold it. The moment a number has to stay reachable, the country stops being a price and starts being a commitment.

Two things shift with it. The first is that the country now has to keep working for the whole term rather than for one lookup, so a market with several routes and steady stock is worth more than the last few cents of savings. The second is that anything the account learns from the number — its region, its dialling code, the country a future re-check will expect — is now attached to something you will keep using. Picking the cheapest market for a 90-day line means living in that country's number for 90 days.

The practical rule is to rent where you would be content to be re-verified. If the account is one you use from Europe, a European line will look ordinary every time it is checked; a line from a market you have no other connection to is the kind of mismatch some platforms score against you. Rental prices run per country and per tier, so the comparison to make is between markets that would each be fine, not between the cheapest and the rest.

The country you cannot change later

Whatever country you verify from becomes the number on the account, and on most platforms changing it means running the whole verification again on a new line. With a one-time code that is not much of a loss — the original number was gone twenty minutes after you used it anyway. It becomes a real cost when the account has since accumulated something worth keeping, or when the platform treats a change of number as a security event and locks things down while it thinks about it.

So the decision to take slowly is the one for accounts you intend to keep, and the honest answer there is often not a country at all. Put an authenticator app on the account and save its backup codes: it never expires, it does not depend on anyone holding a line, and it is stronger than SMS on its own merits. Keep the bought number for the step that demanded a phone, and do not build a long-term second factor on a line that returns to the pool the same afternoon. If SMS genuinely has to be the second factor, rent the line it runs on.

Choosing a country from a script

The REST API exposes the same catalogue the sidebar draws on: countries, services, live prices and the routes under each, so an ordering script can sort candidates by price, filter by tier and request a number without a human reading the table. A failed activation refunds on cancel, which means the sensible escalation loop — cheapest acceptable tier first, one step up on refusal, next country only after two tiers fail — costs nothing per failed attempt beyond the time.

The mistake to avoid in code is the one people make by hand, only faster: sorting the country list by price and walking it top to bottom. That burns attempts across markets while never testing the variable that actually failed. Sort by price within a country, escalate the tier, and change country only when the country is what is left. Autonomous agents reach the same catalogue over MCP with the same order lifecycle and the same rule applies.

What this guide will not claim

There is no country that works everywhere, and any list ranking countries by success rate is describing last month at best. Acceptance is decided by each platform against each range and it moves — a market that clears everything in January can be the one being filtered by March. What is stable enough to plan around is the structure: the tier is read before the SMS is sent, the country is read by the platform's own coverage rules, and those two facts do not change with the season.

Nor can we promise a country will be accepted. What is promised is narrower and checkable: a live line on the route you chose, the code inside the 20-minute window, and the full price back automatically when nothing arrives — no ticket, no claim. That refund is what makes escalating a tier cheap enough to be the default move, and it is the reason the advice here is to test one variable at a time rather than to buy the most expensive thing on the shelf.

And a number from another country is not a location. It changes what a signup form is told and nothing else — not your address, not your device, not what the platform records about how you use the account. Choosing a distant country because it sounds more private mostly buys you a mismatch that a fraud model can see. What we keep, and for how long is a separate question we answer in full, and the Monero guide covers the payment half of the same chain.