What's new in Artisse

Artisse has changed substantially since the last product update posted here. If you used it a while ago, or you are arriving from an older article, this is what the app is now.

Templates instead of prompts

The central change. You no longer describe a photo you want in text and hope. You pick a template — a finished photographic setup where the lighting, the setting, the framing and the styling are already decided — and your face is rendered into it.

The reason is repeatability. A written description produces a distribution of results, so two goes at the same idea give you two different photographs and a lot of retries. A template holds every variable fixed except you, which is why a set generated from one template looks like it came from a single session.

Five categories, and over two hundred templates

The catalogue is organised into five categories, each doing a genuinely different job:

• Flash Snap — hard direct flash at night, the compact-camera look.
• Vibe Snap — strong colour and mood, closer to a film still.
• Profile — close crops built to stay legible at small sizes.
• Daily — natural everyday photos in ordinary places. The largest category.
• Professional — neutral backgrounds and business styling, for headshots.

New templates are added regularly, and the catalogue is ordered by region, so the ones most used where you are appear first. Every template is available to everyone; only the order differs.

You can browse before signing up

The whole public catalogue is open to guests. You can look through every template, in either the app or the browser, without an account — and decide whether it is worth your time before handing over an email address. Generating still needs an account, because it needs your face model.

The full catalogue is also browsable here on the site.

iPhone, Android and the web

All three are live. Earlier versions of this site said iOS and Android were coming soon; they are not coming, they are here, and there is a browser version as well. Your face model and your photos are the same account across all three.

How it works now, end to end

1. Upload around ten ordinary selfies, once. Variety in lighting and angle matters far more than how good the photos are — there is a short guide to that here.
2. Pick a template from any category.
3. Generate. Roughly three minutes per photo, at full resolution, and the results are yours to use.

The face model is trained once and reused by every template, so trying a second look does not mean starting again.

Pricing

Credits, either as a one-off pack or through a plan with a monthly credit allowance. Plans are billed monthly or yearly. Current figures for every tier and pack are on the pricing page, which reads them from the same source the app does rather than repeating a number that quietly goes stale.

What has not changed

The constraint the product is built around is the same one it always was: these are generated photographs, not records of events. They are good for how you look and bad for asserting where you were. The guide to AI photography covers where that line sits, including the places a generated photo should not be used at all.