DeepCut EntertainmentSwitzerland
We make entertainment.
- Film
- Photography
- Social
- Talent
- Real × Digital
- AI Research
We build attention, and the systems behind it.
- aAttention is the only genuinely scarce thing we sell.
- bEverything on screen is made in-house — brief to master.
- cAnd seven of the systems are running on this website.
DeepCut is a Swiss entertainment company. We develop talent, direct film and photography, and produce content for the platforms modern audiences actually live on.
And we do something most production companies do not: we run a research practice. DeepCut Labs studies how machines read intent and context, and builds the experimental systems that follow from it. The two halves are not adjacent — the work feeds the research and the research comes back into the work.
- 01
Content
Film, photography and platform-native production.
Concept, direction, shooting, edit and delivery under one roof. A campaign film and the thirty vertical cutdowns it has to survive as are treated as one brief, not as a job and an afterthought.
See the work - 02
Talent
Real performers and digitally developed characters.
We cast, direct and build the people on screen — and, increasingly, the ones who were never in front of a camera at all. Both are held to the same standard of presence, consistency and craft.
Meet the roster - 03
Labs
Applied research into intelligent systems.
A working research practice inside an entertainment company, studying how machines read intent and context — and building the experimental systems that follow from it.
Enter Labs
Made to be looked at twice.
Everything is shot, cut, graded and finished in-house. That is not a boast about facilities — it is the reason a film and the thirty cutdowns it becomes still look like they came from the same company six weeks later.
The cut
Six shots, one frame. Scroll to cut between them.
The aperture changes shape to meet each format, and nothing dissolves — every change is a hard cut on a frame boundary, because that is what an edit is. The range is the point: wide to macro, motion to stillness, 2.4:1 to 9:16, without the register ever slipping.
A photograph has to hold up with nothing next to it.
Which is why the grid gets out of the way here. Scale, negative space and one image are a composition; nine thumbnails are a content management system.
Real and digital, on the same call sheet.
DeepCut works with people in front of cameras and with characters that were never in front of one. The interesting part is not that both exist — it is that the direction, the continuity and the standard have to be identical, or the audience finds the seam immediately.
LiveBoth halves are the same frame. The right one is rebuilt every pass from a coarse luminance and gradient read of the left — drag the divider.
Nothing here is presented as a matched pair of identities. It is one recording, described twice: once as a lens saw it and once as a system measured it. That distinction is the whole discipline.
Faces that make people stop.
Casting is the first and most expensive decision on any production. We keep it in-house — real performers, creators who already carry an audience, and characters developed entirely in our own pipeline. Which of the three a project needs is a creative question, and it gets answered before the budget does.
Portrait — studio
Performance — location
Motion — presence test
The roster is built per project, not sold from a catalogue.
How casting works at DeepCutThe other half of the company.
DeepCut Labs researches how machines read intent and context, and builds the experimental systems that follow from it.
It exists inside a production company on purpose. Every idea has a real brief, a real deadline and a real audience to fail in front of. Some of the work is published. Some of it is in development and stays closed.
- Intent field
- Decision surface
- Anticipation
- Context lattice
- Persistence
- Commit gate
- Redaction
Seven live systems. Not one of them is a video of a system.
Enter LabsA small company that finishes things.
Swiss, independent, and deliberately not large.
DeepCut is built so that the people who take the brief are the people who make the work. There is no account layer between a client and the person holding the standard, which is the only reliable way we have found to keep a thing from getting worse between the pitch and the delivery.
- 01
Direction
The idea, the treatment and the person holding the standard from first conversation to final master.
- 02
Film
Campaign films, brand pieces, fashion and beauty motion, documentary-register interviews and everything cut out of them.
- 03
Photography
Editorial, campaign, portrait and product. Shot to work at billboard scale and in a 4:5 crop on a phone.
- 04
Short-form
Vertical work built for the platform it lands on, not reformatted into it after the fact.
- 05
Post & finishing
Edit, grade, sound, motion and versioning. The pass that decides whether the work looks expensive or looks close.
- 06
Applied AI
Research and internal tooling from DeepCut Labs, used on our own productions before it is used anywhere else.
Vertical is a format, not a compromise.Vertical isa format.Not acompromise.
Most brands still shoot a film and then apologise for it in 9:16. We board the vertical first.
A short-form piece has about six frames to earn the rest of itself. That constraint changes where the camera goes, how the wardrobe reads at thumbnail size, where the caption sits, and whether the loop closes. None of that is a post-production decision, and none of it survives being retrofitted.
One take, four frames
Watch what reformatting actually costs.
This is a single vertical take, and the screen is changing its own shape around it. Nothing is being re-shot or re-encoded — the same footage is simply being asked to survive four different frames, and you can stop it on any of them.
By 16:9 there is almost nothing left. That is the argument for boarding the vertical first, made with the material rather than asserted about it. The housing is ours: no status bar, no borrowed interface, no imitation of a platform.
Three platforms.
Three grammars.
One shoot.
Every frame it has to live in.