Check whether a phone number is VoIP, mobile or landline, and which carrier its number block belongs to. Free, no account, nothing is called or texted.
25 free checks a day from this address.
Run the whole file through the free cleaner — it works in this browser, needs no account, and the file is not uploaded anywhere. That covers formatting, duplicates and numbers that cannot be valid.
For the parts a browser cannot do — whether each block was ever allocated, its line type and its original carrier — we will email a trial key for the desktop app: 50,000 numbers, offline, no card, Windows only. It also raises this page from 25 checks a day to 100. Nothing else is sent to you.
A VoIP number looks exactly like any other number. Nothing in the digits marks it. What differs is who it was issued to: VoIP numbers come from blocks held by internet telephony carriers rather than mobile networks or local phone companies.
This lookup matches the number to its block and reads two things from the numbering-plan record: the line type, and the carrier the block was allocated to. In the US, large VoIP wholesale carriers such as Bandwidth and Onvoy hold blocks in most area codes, and the numbers sold by many calling apps and business phone services come from those blocks. Some blocks are labelled landline in the source data even though they are held by VoIP carriers; Phonelint corrects those using the carrier.
The number was issued from a block held by a VoIP provider, and which provider. For list cleaning and SMS routing, that is usually the question that matters.
Fixed or non-fixed VoIP. A business desk line tied to an address and a free app number anyone can create in a minute can both be VoIP. Telling them apart needs a paid carrier database or a live query.
Ported numbers. A mobile number ported to a VoIP service still reads as mobile here, and a VoIP number ported to a mobile network still reads as VoIP. Portability does not update the block record. A live lookup through Twilio or Telnyx returns the current network.
Who owns it. This tool describes the number, never the person.
There are 394,736 US number blocks allocated to VoIP, held across 307 distinct carrier records. They are not spread evenly. Two wholesalers hold two thirds of them:
| Carrier | VoIP blocks | Share |
|---|---|---|
| Onvoy (Inteliquent) | 182,557 | 46.2% |
| Bandwidth.com | 82,039 | 20.8% |
| Comcast IP Phone | 23,752 | 6.0% |
| Terra Nova Telecom | 13,517 | 3.4% |
| Telnyx | 13,480 | 3.4% |
| CenturyLink Communications | 11,369 | 2.9% |
| Commio | 9,873 | 2.5% |
| IP Horizon | 9,591 | 2.4% |
Counted from Phonelint's numbering table, 1 November 2026. Carrier names are normalised, so "ONVOY, LLC- NY" and "Onvoy, Llc - Ca (Inteliquent)" are counted once.
This explains the result that confuses people most. Onvoy and Bandwidth.com are numbering wholesalers, not apps. Google Voice, support desks, dialers and dozens of messaging platforms rent their numbers from carriers like these, so the allocation record names the wholesaler and not the service the number is used through. A number you know is Google Voice can legitimately come back as Onvoy. The result is correct about who holds the block; it was never a claim about which app answers the call.
VoIP is the hardest of the three line types to get right from allocation data, because a block reassigned to a VoIP provider does not always carry a new classification. Measured against two paid live services, out of sample, method registered beforehand:
| Measured against | Agreement on line type | Sample |
|---|---|---|
| Veriphone | 94% | n = 582 |
| Twilio Lookup | 75% | n = 513 |
| Twilio and Veriphone, with each other | 73% | — |
Benchmark v2, September 2026. Method and raw counts in the accuracy benchmark.
The two paid references agree with each other only 73% of the time, which is the context any single figure needs. There is no oracle for line type: "agreement with Twilio" is a comparison, not a grade, and a service quoting you one accuracy number for VoIP detection is quoting its own opinion as ground truth.
What that means in practice: treat a VoIP result as a strong signal and not a verdict. It is reliable enough to sort a list by risk, and not reliable enough to delete a customer record on.
The usual reason to find VoIP numbers in bulk is deliverability or fraud screening. Both go wrong in the same way: treating 12% of US exchanges as disposable.
VoIP is not a synonym for fake. 27,229 US exchanges are allocated to VoIP carriers, and they include Comcast's residential phone service, every business running a hosted PBX, and most contact centres. Deleting VoIP wholesale removes real customers, and on a B2B list it removes a large share of them.
What VoIP does predict is that a number is cheap to obtain and cheap to discard, which is why it correlates with throwaway signups. That makes it a useful input to a risk score and a bad basis for a blanket rule. Score it; do not filter on it alone.
SMS is the one place to be careful. Delivery to VoIP numbers is inconsistent — some carriers accept and route it, some silently drop it, and the sender is usually billed either way. If an SMS campaign is the use case, segmenting VoIP into its own send and measuring delivery separately tells you more than guessing.
The free bulk cleaner splits a file by line type in the browser without uploading it, which is enough to see how much of a list is VoIP before deciding anything. The desktop app does the same across 1.88 million blocks offline at about 47,000 numbers a second, measured on a mid-range Windows laptop, and can write each line type to its own file.
For outreach lists the more useful split is mobile versus everything else, because that is the line that decides which numbers can take an SMS at all. VoIP sits on the wrong side of it often enough to matter and not reliably enough to trust, which is the practical case for segmenting rather than deleting. The landline or cell checker answers that split for a single number, and the guide to line types covers how the three types behave.
One caveat that applies to every number in the output: line type comes from the block's original allocation, so a number ported to a VoIP service after it was issued still reads as whatever it started as. The line-type guide covers how often that happens.
A lot of people arrive here holding one number from an eSIM or virtual-number service, asking whether it is VoIP, because PayPal, Telegram and some banks refuse VoIP numbers at sign-up. The common move is to ask whether a particular area code is VoIP. It is the wrong question, and the data says so plainly. These are three area codes eSIM services hand out often:
Six-digit exchanges per area code, counted from Phonelint's numbering table. Remainders are special-use or unclassified ranges.
No area code is all one thing. 669 is 42% VoIP and 40% mobile, so two numbers with the same area code can honestly give opposite answers. "Is 934 mobile?" has no answer at the area-code level; the number you actually hold does. That is what the box above is for — it reads the specific six-digit block your number sits in, not the first three digits.
What this cannot promise you. It reports what a number was issued as. It cannot tell you whether a given service will accept it. Sign-up checks at PayPal, Telegram and the rest use live data that includes porting history, carrier reputation and signals nobody publishes, and they each draw the line somewhere different. A number this page calls mobile can still be refused, and a number it calls VoIP is sometimes accepted. Use it to find out what you are holding, not as a prediction about somebody else's fraud rules.
If a provider gave you a number and you would rather have a different one, knowing the line type is the useful part of that conversation: it turns "I do not like this number" into "this one is issued from a VoIP block, and I need one that is not."
Enter it above. If the number's block belongs to a VoIP provider, the result says VoIP and names the carrier. The digits alone cannot tell you; it takes a lookup against numbering-plan data.
Often, yes. Many VoIP services support SMS, but delivery is less consistent than on mobile networks, and some short-code and verification senders do not deliver to VoIP numbers at all.
No. Both come from VoIP carriers' blocks. Separating them needs a paid carrier database or a live query, which free tools do not make.
Some app numbers are ported in from mobile or landline carriers, and a port does not change the block record. Only a live lookup returns the network a ported number is on today.
Yes, with no account, up to 25 checks a day, or 100 a day if you add an email address.
Yes. The Phonelint desktop app labels every number in a list as mobile, landline or VoIP offline, without uploading it. The free bulk cleaner handles formatting and duplicates in the browser.
Enter it in the box above. eSIM and virtual-number services hand out numbers from whatever blocks they hold, and those vary within a single area code — 669 is 42% VoIP and 40% mobile, so the area code tells you nothing on its own. The result reads the specific six-digit block your number sits in.
None of them is, as a whole. Counted across their exchanges, 669 is 42% VoIP and 40% mobile, 934 is 38% VoIP and 34% mobile, and 940 is 16% VoIP against 36% landline. Two numbers sharing an area code can return opposite answers, so the only reliable check is the number itself.
This cannot tell you that, and no free tool can. It reports what the number was issued as. Sign-up checks use live data including porting history and signals that are not published, and each service sets its own threshold. A number shown here as mobile can still be refused.
Clean 50,000 numbers in your browser free, or label millions by line type offline with the desktop app. Neither uploads your list.