An empty post-production suite at night, lit by a single violet lamp

About LumarLabs

This room used to be
the price of entry.

A suite like it, or an afternoon of someone who owned one. LumarLabs exists because that stopped being a reasonable thing to ask of a person with a good idea and a camera.

What we do

We make four things: a studio for swapping a face into video and photos, an avatar that reads a script aloud, a live swap that holds up on a video call, and a voice you can wear while you speak.

None of that is new on its own. What was missing was a version of it a person could actually reach — one that did not open with a sales call, did not need a machine you had to specify in advance, and did not price itself for a company when the person using it is one person.

So the whole product is built around the assumption that you want to try it before you believe any of this. You can, on your own footage, without giving us a card.

The same job, either side of a line.

What it took

  • A suite you booked by the day, or someone who owned one.
  • A quote first, and a contract before you could test whether it worked.
  • Changes measured in days, because they went back through the editor.
  • A finished file you could not touch again without paying to reopen it.

What it takes

  • A laptop, and an account you made two minutes ago.
  • You try it before you decide anything, on your own footage.
  • Changes are another render. You make them yourself.
  • Your files stay in your library. Re-run them, replace them, delete them.

Four things we will not trade away.

Training

What you upload does not train anything.

Your footage, your faces and the videos you make are yours. They are not added to a training set, not used to improve a model, and not shared with anyone else. This is in the privacy policy in the same words, so you can hold us to it.

Pricing

The price you see is one you can actually pay.

Prices show in your own currency only where there is a way to take it. Quoting a figure in a currency we cannot accept is worse than quoting nothing — you would find out at checkout, which is the wrong moment.

Reach

One account, whichever screen is nearest.

The web app, the browser extension, the desktop app for Windows and macOS, and iPhone and Android all sign in to the same account and the same library. Start something on a phone, finish it at a desk, and nothing has to be moved across by hand.

Processing

We tell you where your video is processed.

Some of what we do runs on your machine and some of it runs on ours, and the difference matters enough that it gets its own section on this page rather than a footnote in a policy.

Where your video goes

Some of this runs on your machine. Most of it runs on ours.

Products in this category tend to say “on-device” and leave it there, because it is the answer people want. Only one of the things we make is genuinely on-device, so here is the whole list instead.

Live Swap, in the desktop app

Your machine

The model downloads once, the first time you open the app. After that every frame is processed on your own computer and your camera feed does not reach us.

Live Swap, in the browser extension

Our servers

The extension carries no model — a browser extension is not allowed to fetch one — so the swap happens on our side. Same feature, same calls, different place the work is done.

Live Swap Pro, everywhere it runs

Our servers

This one is too heavy to run on a laptop while a call is going, on any platform. Your camera goes up, comes back changed, and streams into the call. It is not the on-device path and we will not describe it as one.

Face Studio, Avatar, Lip Sync, Identity Swap

Our servers

Renders run on our hardware, which is why they finish faster than your laptop would manage. Your files are encrypted on the way there and while they sit with us, and you can delete any of them from your library.

Wherever it runs, what you upload is not used to train models and is not sold or shared. The detail is in the privacy policy.

The quickest way to judge this is to make something.

There is a free tier, and it does not ask for a card. Bring a clip you already have and see whether the result is one you would actually use.