VoIP vs Non-VoIP Numbers: Why One Verifies and One Fails
You have two phone numbers. Both make calls. Both receive texts. You type the first one into a signup form and get a code within seconds. You type the second one in and the form rejects it before it even tries. The VoIP vs non-VoIP distinction is the entire explanation, and it has almost nothing to do with what the numbers can do and everything to do with how they were created.
This is a comparison rather than a definition. If you want the ground-up version first, read what a non-VoIP number is. What follows is the head-to-head: how the two types differ technically, what each costs, which one a given platform will take, and — the part most articles skip — when VoIP is genuinely the better choice.
VoIP vs non-VoIP: what actually differs under the hood
A non-VoIP number is assigned by a licensed carrier and attached to a subscriber record on that carrier's network. If it's a mobile number, there's a SIM (or eSIM profile) somewhere that the network can query. Messages to it travel over the carrier's own signalling infrastructure and terminate on a device the network knows about.
A VoIP number is assigned out of a block that a provider leases, and it exists as a routing rule in software. When a message arrives, the provider looks up where to deliver it — an app, a webhook, an email, a SIP endpoint — and forwards it over the internet. There is no SIM and no subscriber record on a mobile network, because there is no mobile network involved.
Both numbers look identical when written down. The difference is entirely in the records behind them, and those records are public enough that any website can read them in a fraction of a second.
| Non-VoIP (mobile) | VoIP (non-fixed) | |
|---|---|---|
| Who issues it | Licensed mobile carrier | Software provider leasing number blocks |
| Physical requirement | SIM or eSIM on a real network | None — an internet connection |
| Time to obtain | Minutes to days; often ID required | Seconds, via API, at any volume |
| How SMS reaches you | Carrier signalling to a handset | Internet delivery to an app or endpoint |
| Line type on lookup | mobile | nonFixedVoip |
| Typical cost | Carrier plan, or per-verification from a provider | Roughly $1–2/month plus fractions of a cent per message |
| Verification outcome | Usually accepted | Frequently rejected before a code is sent |
| Genuinely good for | Account verification, KYC, anything identity-linked | Business calling, routing, teams, software integrations |
The three-cent check that decides everything
People imagine platforms are guessing. They aren't. Line-type detection is a commodity API call, and it is cheap enough that there's no reason not to run it on every signup.
Twilio's Lookup API is the reference example. Basic formatting and validation is free. The Line Type Intelligence package — which returns whether a number is mobile, landline, fixed VoIP, non-fixed VoIP or toll-free, plus the carrier name — is listed at $0.008 per request. Line Status, which checks whether a number is currently reachable on a network, runs about $0.007 per request, falling to roughly $0.00385 at volume. Vonage, Telesign, Telnyx and IPQualityScore all sell comparable products.
At those prices, screening a million signups for line type costs a few thousand dollars. For any platform with a fraud problem, that is trivially worth it.
There are two flavours of check, and knowing which one you've hit explains different failure modes:
- Database lookup (LRN-based). Reads industry routing and number-portability records to classify the number. Instant, static, and works whether or not the number is currently switched on. This is what catches VoIP.
- HLR lookup (live network query). Sends a request to the mobile network itself and comes back with the current carrier, whether the SIM is active, and sometimes roaming status. This is what catches a number that is technically mobile but disconnected, or one whose SIM isn't live.
A VoIP number fails the first check and simply doesn't exist as far as the second is concerned — there's no home network to interrogate. That combination is why "port it to look like a mobile" advice rarely works: portability records generally follow a number, and an HLR query on it still returns nothing.
Which platforms actually care
Policies vary a lot, and they change without announcement. The broad pattern is that the more a service ties your phone number to money or identity, the less tolerance it has.
| Category | Examples | Typical strictness with VoIP |
|---|---|---|
| Messaging apps | WhatsApp, Telegram, Signal | Strict. WhatsApp's guidance states VoIP numbers aren't supported for registration, along with toll-free, premium-rate and universal-access numbers. |
| Banks, payments, crypto | PayPal, Coinbase, retail banks | Strictest. Phone type feeds directly into KYC and fraud scoring. |
| Major account providers | Google, Apple, Microsoft | Strict and inconsistent. Often depends on how new the account is and where you're signing in from. |
| Social platforms | Instagram, X, Discord, Snapchat | Moderate. Frequently accepted at signup, then rejected later when re-verification is triggered. |
| Marketplaces and gig apps | Amazon, eBay, Uber, DoorDash | Moderate to strict, especially for seller or driver accounts. |
| Smaller sites and tools | Newsletters, SaaS trials, forums | Usually don't check at all. |
The uneven behaviour in that middle band is what confuses people most. A number that worked fine for six months suddenly stops working when a platform runs a re-verification sweep or tightens a threshold. Nothing about your number changed; the rule did.
Where VoIP is genuinely the better choice
Most articles on this topic treat VoIP as the villain. It isn't. It's the wrong tool for one specific job — proving identity to a third party — and the right tool for a lot of others.
- Business phone systems. Ring groups, IVR menus, call queues, voicemail-to-email, hunt sequences. Carrier mobile service does none of this well; VoIP does all of it cheaply.
- Distributed teams. One number, answered by whoever's on shift, from whatever device they have. That's a VoIP feature, not a carrier one.
- A public-facing contact number. Putting a VoIP number on your website keeps your personal mobile off the open internet, and you can change where it routes without reprinting anything.
- International calling. Per-minute VoIP rates are usually a fraction of carrier roaming or international dialling.
- Software integrations. Appointment reminders, delivery notifications, order updates, two-way support messaging — these are all built on VoIP-style programmable numbers, and it would be absurd to run them on SIMs.
The practical answer for most people is that both types have a place: a carrier-backed number for anything that has to survive a verification check, and VoIP for the day-to-day communication work it's actually designed for.
Where non-VoIP is the only thing that works
Conversely, some situations leave you no option:
- Registering a WhatsApp or WhatsApp Business account. The policy here is explicit, not a soft preference.
- Financial and crypto onboarding. Phone line type is an input to the risk model, so a VoIP number can cause friction even after signup succeeds.
- Verifying with a US-only service from abroad. You need a real US mobile line type, not just a US area code — a VoIP number with a 212 prefix is still VoIP.
- Keeping your personal number private on genuine signups. The privacy motive is legitimate; the filter just doesn't distinguish it from anything else, so the number you use still has to be carrier-backed.
- Business or vendor registrations that explicitly require a non-VoIP contact line.
How to tell which one you're holding
- Consider how you got it. If it arrived in under a minute with no SIM and no ID, it's non-fixed VoIP. This is right roughly all of the time.
- Run a free line-type lookup. Several providers offer a small free tier. Enter the number in full E.164 format (
+15551234567) and read the type field. - Read the carrier name, not just the type. A recognisable mobile network is a good sign. Names like Bandwidth, Twilio, Telnyx or "Google (Grand Central)" mean VoIP regardless of what else the result says.
- Note whether the rejection was instant. Rejected before any SMS was sent means a line-type filter. Accepted but no code arriving is a different problem — a delivery or routing issue, not a classification one.
The cost comparison, honestly
VoIP is cheap because software is cheap. A US local number from a programmable provider runs about $1.15 per month, with SMS around $0.0083 in each direction plus carrier surcharges of roughly a quarter to half a cent per message. Toll-free is around $2.15 per month.
Carrier-backed non-VoIP numbers cost meaningfully more, because someone had to acquire a real SIM on a real network. Verification services typically price per successful code rather than per month, and the price varies by platform — the harder a service is to verify, the more a working number for it costs.
That gap is not a markup for its own sake. The cost is the signal. A number that's expensive and slow to obtain is exactly what makes it useful as a fraud speed bump, which is the whole reason platforms accept one type and refuse the other. If non-VoIP numbers were free, they'd be blocked too.
One caveat worth stating plainly: be wary of free public numbers advertised for verification. Their inboxes are shared and readable by anyone, and they're usually already registered on every major platform, so they fail on both privacy and functionality.
Frequently asked questions
Is fixed VoIP treated the same as non-fixed VoIP?
Not always. Fixed VoIP is tied to a verified physical address — a business phone system or an ISP-provided home line — and some platforms accept it where they'd refuse non-fixed VoIP. Others block both. It's the least predictable category of the four.
Can I make a VoIP number pass as non-VoIP?
No, and it's worth being clear about why. Line type isn't a label attached to the number that can be edited; it reflects how the number is provisioned and routed. Porting between providers doesn't reset portability records, and an HLR query on a VoIP number returns nothing because there's no network subscriber record to find.
Why did the same number work last year and not now?
Platforms tighten thresholds over time and periodically re-verify existing accounts. A number accepted under a looser rule can fail a later check without anything about the number changing.
Do I need to give up VoIP entirely?
No. Keep VoIP for what it's good at — call routing, business lines, cheap international dialling, software notifications — and use a carrier-backed number for verification. Using one number for both jobs is what causes the problem.
The short version
VoIP and non-VoIP numbers do the same visible things and are built completely differently underneath. Websites read that difference from routing databases for less than a cent, and they act on it because internet numbers are too cheap to be worth anything as proof of identity. Choose based on the job: VoIP for communicating, carrier-backed for verifying. If a signup keeps refusing your number, line type is nearly always the whole story.