The best video chat apps, judged by seven things you can test yourself
No ranking, no affiliate list, no scores somebody made up. Seven checks you can run in about a minute each, and further down, an honest account of where this site passes them and where it doesn't.
Start video chatWhoChat Live is one of the products being measured here, which is a thing you should factor in. That's why every test below is one you can run without us.
People you could be matched with
01 / THE PROBLEM
Why a ranked list of video chat apps is mostly useless
Look at how one of those lists is built. A number appears beside each product: 9.2, four and a half stars, "our pick", and nowhere on the page is there a word about what was measured, on which device, by whom, or on what day. Often nobody opened the thing at all. The order is just the order. You're asked to take it.
The deeper problem is that "best" isn't a property of the product. It's a property of the fit between the product and one specific person's evening. Somebody who wants to filter for a particular kind of stranger and somebody who wants a face on screen within three seconds are shopping for opposite things, and any single ordering is going to be wrong for one of them.
So this page ranks nothing. What it hands you instead is seven checks that genuinely separate these products, every one of them runnable by you, on whatever device is already in front of you, in about a minute apiece, with no expertise and nobody's word to take. Then it puts us through the same seven in public. Losses first.
02 / THE SEVEN
Seven things you can test in a minute each
None of these is an opinion. Every one of them produces a number, a yes, or a no, which is the entire point. You shouldn't have to trust anybody, including us.
- 01
Seconds to the first face
Do this: Open it and count. Not "is it fast": an actual number, from the moment you press the button to the moment a person is on your screen.
Everything in this category lives or dies on this number. Somewhere past four or five seconds people stop waiting and go and do something else instead, and the product never gets a second chance to show them anything.
- 02
What it wants before that
Do this: Count the steps standing between opening it and talking: install, account, email, phone number, date of birth, a profile photo, five interests from a grid.
Each one loses a share of the people who were willing a second ago. A phone number is the heaviest of them by a distance, and it is the one worth being most suspicious of, because a product that needs your number before it shows you anything has a plan for it.
- 03
Where the money sits
Do this: Find the wall before you hit it. Is it in front of the first conversation, in the middle of one, or on a filter you did not know you wanted?
This is the difference between a product that charges and a product that ambushes. A wall you can see from the outside is fine. A wall that appears ninety seconds into a conversation, while somebody is mid-sentence, is a design decision about you rather than about the price.
- 04
Whether the people are people
Do this: Ask something only a live human in that room could answer. What is behind you. What the weather is doing. Wave and see if they wave back within a beat.
Recorded loops and scripted accounts are the oldest trick in this category and they are still running. The tell is timing: a real person answers a stupid question in about a second, and a loop keeps doing whatever it was already doing.
- 05
How hard it is to report somebody
Do this: Find the report control before you need it. Count the taps. See whether it makes you type an explanation.
A report that asks you to describe what happened gets abandoned, because the thing you just saw is the thing you least want to write out. Buried three menus deep is the same as absent, and the products that mean it put the button where your thumb already is.
- 06
What it keeps after you leave
Do this: Look for three things: is the video recorded, is there a profile, is there a history you could go back and read.
Every one of those is a thing that can leak, be subpoenaed, or be sold. The safest version of this is the boring one where nothing was written down in the first place.
- 07
Whether it runs on what you are holding
Do this: Try it on the device that is actually in front of you right now, not the one the review used.
Most of this category is phone-first and a good share of it is phone-only. If you are at a desk, that is the difference between trying something and adding it to a list of things you will never get round to.
03 / OUR OWN SCORE
Where we land on our own list
Losses first, because a page that grades itself and leads with the wins isn't worth reading and you already know that.
- Loss · pickingYou cannot pick who you get. The pairing itself runs random. You press start and take whoever comes up.
- Loss · free tierThe free part has an end. You start with a batch of matches and a run of free talk time; past that, video pairing runs on coins.
- Loss · app featuresAR masks, beauty filters, that whole shelf of things to put on your face. It belongs to installed apps and we are not one.
- Seconds to first faceThe search starts while the camera permission dialog is still open, so the wait is usually under a second once you say yes.
- What we ask firstThe camera permission. That is the list. A guest session is created in the background: no email, no number, no profile.
- ReportingOne tap, no typing. The session goes with the report, so you never have to describe or prove anything.
- What we keepThe stream is not recorded or stored, and no profile is created, so there is nothing to delete afterwards.
- DeviceA browser tab. Laptop, desktop, phone, tablet, whatever already has a camera.
Read that top block again before the bottom one talks you round. If picking who you get is the thing you actually want, no amount of speed further down the list makes up for it, and you should go and use something that sells you a filter.
04 / THE UNIT
"App" is a strange unit to shop in
The search everybody types has the word app in it, and then a good share of what they end up using is a web page. The word is doing a job it wasn't built for: it means "the thing I use", not "the thing in a store".
Which unit is better depends entirely on what you're after. Installed software can wake your phone up, put masks on your face and keep running in the background, and that's a real set of advantages nobody should pretend away. A page can be opened on anything with a camera, needs no store account, survives on a work laptop, and leaves when you close it.
If you want the two put side by side properly rather than in three sentences, there is a longer comparison here, and a second one against a product that does both: an app store listing and a web version at the same time. And if the format itself is the part you're unsure about (one room, two people), that's its own page.
05 / LIMITS
What a minute of testing can't tell you
Three things, and they happen to be the three that decide whether you're still using something a month later.
Whether moderation actually acts. You can time how long it takes to file a report. You can't see what happened to it. The only honest way to learn that is to report somebody and then keep using the product for a few weeks to find out whether the same faces keep turning up, which isn't a test anybody is going to run on a Tuesday evening before choosing where to spend an hour.
Who is on it at the hour you use it. A product tested at lunchtime and the same product at two in the morning are close to different services: different countries awake, different reasons for being there, conversations that run a different length. One session tells you almost nothing about this. Three sessions at three different hours tells you most of it.
Whether it stays this way. Free tiers shrink. Filters that were free get sold. A report button that was one tap becomes a form. Nothing you measure today is a promise about next quarter, from us or from anybody else, and the only defence is that the tests are cheap enough to re-run whenever something feels like it has changed.
06 / RED FLAGS
Four things worth closing a tab over
- It wants your phone number first. Before it has shown you a single thing. A number is worth money to somebody, and asking for it up front tells you who that somebody is.
- The payment screen arrives before the first face. Charging is fine. Charging you to find out whether there is anything there is not.
- You can't find the report control. Go and look for it now, while nothing is happening. If it takes you more than a few seconds to find, it won't exist at the moment you need it.
- Nobody reacts to you in real time. Wave. Say something specific about what's behind them. If the timing is off by a beat, you're watching a recording of a person, not a person.
None of those four requires any expertise to spot and all four can be checked before you've invested anything but a minute. That's the whole argument of this page: the checks that matter are cheap, so there's no good reason to outsource them to a list that was written by somebody who never opened the product.
07 / QUESTIONS
Questions people ask before picking one
What makes a video chat app good?
Seven things you can check yourself in about a minute: how many seconds until a real face appears, how much it demands before that, where the paywall sits, whether the people are live, how quickly you can report somebody, what it keeps after you close it, and whether it runs on the device you already have in front of you. Everything else on a typical list is decoration.
Which is the best video chat app?
There is no single answer, and anybody giving you one is usually being paid for it. The honest version is that the answer depends on which of the seven things above you care most about. Somebody who wants to filter by gender and somebody who wants to be talking inside three seconds are shopping for different products, and no ranking can serve both.
Are free video chat apps actually free?
Usually free to start and rarely free forever. The useful question is not whether money appears but where. A price you can see before you begin is a different thing from one that surfaces mid-conversation. Here, to be specific about our own: opening the site and your first chats cost nothing, and after the free matches and free talk time are used up, video pairing runs on coins you can buy or collect free from the reward centre.
Do I have to download an app?
Not for this one. It runs in a browser tab on a phone or a laptop, with nothing to install and nothing to sign up for. Plenty of good products in this category are app-only, though, and that is not a knock on them. If what you want is push notifications pulling you back tomorrow evening and a shelf of camera effects to put on your face, an installed app is going to serve you properly and this will not.
How do I know if I am talking to a real person?
Ask something only a live human in that room could answer, then watch the timing. A person replies to a daft question in about a second. A recorded loop carries on doing whatever it was already doing, and once you have seen that gap you never fail to notice it again.
Which video chat apps work on a laptop?
Fewer than you would expect. The category is overwhelmingly phone-first, and a fair number of the best-known names have no desktop path at all. Anything that runs in a browser works on a laptop by definition, which is one of the quieter reasons people end up on browser-based ones.
Is random video chat safe?
It is as safe as its reporting path and its age rules, and no safer. Before you rely on one, find the report button and check what it costs you to use it. Ours is one tap and asks you to write nothing. It is 18+, nudity ends the session, and no, that does not make it perfect. Anyone claiming perfect moderation in this category is selling something.
What should make me close a video chat product immediately?
A phone number demanded before anything is shown. A payment screen in front of the first conversation. No visible way to report. And faces that do not react to you in real time. Any one of those is enough on its own.
08 / RUN IT
Run the first test on this one
Press the button and count the seconds until a face appears. That's test one, and it costs you nothing to find out how we do on it.
Start video chat