China Stopped Copying Car Design. Now the West Is Studying Chengdu.
For most of the last decade, the story Western car studios told about China was a defensive one: worry about being copied. Chengdu — one of China's "big four" motor shows and a launchpad for new-energy-vehicle reveals — was where you went to spot the imitations. The Chengdu Motor Show 2026, opening 21 August, confirms how completely that story has flipped. The copying era is over. Design leadership has moved East — and, more precisely, it moved upstream, to the concept phase where a brand decides what it wants to be.
Look at who is now doing the deciding. BYD, the world's largest maker of electric vehicles, runs its design under Wolfgang Egger, the former head of design for Audi and Alfa Romeo. NIO) built its design language with Kris Tomasson, who came from BMW and Ford. The talent that once flowed out of Europe's studios as cautionary tales now flows into China as signed contracts. The most ambitious Western designers are taking the China job, because that is where the conviction — and the budget, and the speed — now lives.
Conviction is the right word, because what separates the new Chinese design from the old is not surfacing skill. It is the willingness to commit. Xiaomi's SU7 — a phone company's first car — arrived with a face and a stance so resolved that critics accused it of borrowing from Porsche. The more useful observation is that it looked like a decision, not a committee compromise. XPeng, Li Auto and Zeekr each showed up with an identity you can name from fifty metres — a legibility that many legacy brands, hedging nervously between the combustion era and the electric one, have quietly lost.
That legibility is a concept-phase achievement, not a styling flourish. A coherent car is not one with beautiful surfaces; it is one where every surface is downstream of a single, early, defended idea about what the object is for. The Chinese EV brands had an accidental advantage here: with no heritage grille to protect and no eighty-year-old silhouette to honour, they were forced to answer the hardest question in design — what should this even be? — from a blank sheet. They treated that blankness as freedom rather than fear, and they answered it fast, at a cadence Western OEMs still struggle to match.
The West is right to be studying Chengdu, but it should study the correct thing. The lesson is not "add more screens" or "copy the split headlights." The lesson is that identity is a decision made before the clay, and that decisions made with conviction read as design and decisions made by consensus read as noise. Geely-owned Zeekr, Denza, and the premium sub-brands multiplying across the sector are not winning because they out-render anyone. They are winning because they decided who they were and then held the line — the exact discipline that produced the great European marques when they were young and hungry.
There is a warning in it too. Conviction can curdle into sameness: a show floor full of slippery, screen-lined, light-bar-fronted EVs can start to blur, and China's own critics have begun to say so. The next competitive edge will not be deciding faster — everyone can do that now — but deciding better: knowing which of a thousand coherent directions is the right one for a specific brand, and having the judgement to reject the other 999. That judgement is the scarce resource, and it cannot be rendered. It has to be made.
Which is the whole point. The tools to generate a thousand resolved-looking concepts are now universal and cheap; a studio in Chengdu and a studio in Turin start the morning with the same generative firepower. What they do not share is the judgement about which concept deserves to become a car. The center of gravity in car design has moved to the concept phase — to the room where someone looks at the options and chooses. China got there first not because it drew better, but because it was willing to decide. At Depix, that room — the one upstream of the surfaces, where identity is settled before a single panel is modelled — is the only one we think is worth building tools for.
Sources:
