For nine days, the film played to almost nobody. It had opened across 245 screens in China, a real release in real cinemas, and in that opening stretch it took the equivalent of roughly a thousand dollars. The ticketing data puts the audience at fewer than three hundred people nationally. On one of those days it made about $137. By the middle of the second week, the number of venues still bothering to run it had collapsed to four.
Then, over the following weeks, that same film took somewhere around nine million dollars.
Nothing was recut. No distributor stepped in with a marketing budget, because there was no marketing budget to begin with and no posters beyond the one. What changed was that the internet found the animation and decided it was funny. Clips of stiff-limbed cattle and strangely lit landscapes started circulating as a joke, and people began buying tickets specifically to see how bad it was. Some of them, by several accounts, drove for hours to get to a screening.
That is the part of this story people keep getting wrong when they retell it. The audience did not show up because someone persuaded them the film was good. They showed up to mock it, and then a lot of them sat there and were unexpectedly moved, and the two things never stopped being true at the same time.
The film is called Niu Lai. It runs 86 minutes. It was made by two people, and by two people I mean two people, not a small team with two credited leads. Xin Yumeng directed it, built the models, animated it, edited it and voiced characters in it. His mother, Sun Lifang, wrote the script, voiced characters as well, and composed the song that plays over the end. Between them they were the crew, the cast and the post-production department. It was the first film either of them had made.
They worked out of Dalian, in the north-east of the country, on second-hand equipment bought online. Xin was in his mid-thirties by the time it came out, and his actual profession is landscape design. That single detail explains more about how the film looks than any amount of criticism does, because the software he built it in is SketchUp.
If you have never used it, SketchUp is a 3D modelling program for architects and designers. It is the tool you reach for when you want to show a client what a courtyard will look like with the trees in, or how a roofline sits against a hillside. It is genuinely good at that. What it is not built for is character animation. It has no proper rigging system of the kind animators rely on, no sophisticated deformation, none of the layered control that lets a professional make a shoulder roll convincingly into a step. Animating a living creature in it is a bit like building a wooden chair with a bread knife. You can do it. Everyone will be able to tell.
So the stiffness that the internet laughed at was not incompetence exactly. It was a structural consequence of the tool. Limbs move as fairly rigid pieces because the software has no elegant way to make them do otherwise. Faces read as slightly frozen because there is no flexible facial rig underneath them. A trained animator watching Niu Lai can more or less reverse-engineer the constraints Xin was working inside, frame by frame, and see the workarounds.
It is worth spelling out what doing every job actually means in animation, because the phrase gets used loosely. On a normal feature, the work is split across departments that barely speak the same language. Someone writes. Someone else designs characters. Modellers build them, riggers construct the invisible skeletons that let them move, animators move them, a lighting team decides where the sun is, a compositor assembles the layers, editors cut, sound designers build every footstep, a composer scores it, and a cast of actors records the dialogue in a booth with a director listening. Each of those is a career. On this film, two people covered all of it. That means the person deciding how a scene should feel emotionally was also the person who had to go and physically move the calf’s front leg, and then later the person who mixed the sound of it landing, and then the person who cut the shot. There is no one to hand anything to and no one to tell you it is not working yet.
He did it for more than five years. The idea dates to around 2021, and the two of them ground at it across half a decade of evenings and weekends and days that nobody was paying for. The reported production cost is startlingly small, in the region of a couple of hundred dollars, which is really a statement about how little cash changed hands rather than a real accounting of what it cost them. What it cost them was five years.
The film itself is a children’s animal story, and it is worth saying plainly that it is not an ironic object or a piece of provocation. A newborn calf goes on a long, dreamlike journey. There is a skylark. There are other animals, some of them alone in the world. The themes are the ones you would expect from a mother and son writing for children: courage, endurance, love, the willingness to give something up for somebody else. One critic writing about the phenomenon described the film as a kind of strange performance art, and suggested the audience reaction was partly a rebellion against polished taste, which is a sharp reading. But the film is not performing anything. It is sincere the whole way through. That turned out to matter enormously.
The poster did not help. The marketing image is a handsome ink-wash illustration, the kind of restrained, traditional-looking artwork that signals a serious animated feature with money behind it. Then the film starts, and the visual language is blocky 3D from an architectural renderer. That gap between what the poster promises and what the screen delivers is the exact gap that internet comedy lives in, and it is a large part of why the clips travelled so fast.
Now the mechanics of the turnaround, which is the genuinely interesting part and the part most coverage skips.
Chinese cinema exhibition responds to demand faster than almost any other market on earth, because screen allocation is reassigned daily rather than locked into weekly contracts. A film that is not selling loses its slots within days. A film that suddenly starts selling can be given slots within hours. That flexibility is usually invisible, because it mostly functions as a machine for quietly euthanising underperformers. Niu Lai is what it looks like running in reverse.
Follow the numbers as the tracking data has them. On one day in the middle of August the film took the equivalent of about $137. Three days later it took the equivalent of roughly $906,000 in a single day. Daily screening counts across the country went from around twenty to several hundred inside one twenty-four-hour period, and at the peak of the surge the film was reportedly being screened something like twenty-two thousand times a day. From four venues to that, in under a fortnight.
No committee decided this. Thousands of individual cinema managers looked at overnight presales, saw a title they had already written off filling seats, and reassigned rooms to it. That is the whole mechanism. It is not a story about a marketing department being clever. It is a story about a distribution system with no opinion of its own, doing what the ticket sales told it to do.
Within about a month, the film had sold well over two million tickets and its gross had passed the equivalent of roughly nine million dollars.
It is also worth being honest about what those first nine days would have felt like. Anyone who has released anything can imagine the specific dread of it. The tracking platforms in China publish box-office data publicly and update it constantly, which means that during those nine days the two of them could watch, hour by hour, the precise numerical measurement of nobody caring. Not a bad review, which at least means someone watched. Not an argument. Just a figure that barely moved, next to the venue count falling, because a film with empty rooms loses its rooms. Five years of work, and the public verdict arrived as a rounding error. Whatever they thought the risk of making this film was, that was the version of failure that actually turned up.
There is a number in this story that you will see reported two different ways, and the difference is not small, so it is worth being straight about it. The opening gross is recorded in the ticketing data as around 7,169 yuan across the first nine days, which is roughly a thousand dollars, and one prominent piece of commentary put that opening figure at seventy-one dollars instead. The two are not reconcilable, and only the first traces back to the box-office tracking platform that the trade reporting cites. That is the figure used above. The lower one has travelled widely because it makes a better headline, which is exactly the reason to be careful with it.
What the reliable version of the number tells you is already extraordinary enough. A thousand dollars of tickets in nine days, from a release across 245 screens, is not a disappointing opening. It is a total absence of an audience. Two hundred and thirty-odd people in a country of over a billion.
The argument that ran through a lot of the commentary afterwards was that viewers had grown tired of frictionless, expensive, machine-smoothed animation, and were responding to something visibly made by hand. There is something in that, and it is worth taking seriously rather than dismissing as sentimentality. A great deal of what people scroll past now is output with no author behind it, generated at volume, competent in a way that means nothing. Niu Lai is the opposite kind of object. Every awkward movement in it is the trace of a decision by a specific person who did not know an easier way. You can see the labour. The critic quoted most often on this called what the film has a kind of pure enthusiasm, uncontaminated by automation, and that is close to the truth of why people stayed in their seats.
But it would be dishonest to turn this into a warm story about the internet choosing kindness. The mockery did not stop. It never became affection for most of the people involved in spreading it. Comparisons to the famous cult disasters of cinema, the films people attend in order to enjoy how wrong they are, ran right through the surge and kept running after it. A large share of that nine million dollars was paid by people who wanted to laugh, and who did laugh, and who posted about laughing on the way home. The film was a national joke and a hit simultaneously, and the second thing happened because of the first.
Sit with that for a second from the point of view of the two people who made it. For nine days, five years of work is invisible. Nobody has seen it. The cinemas are dropping it. Then, inside about a week, the work is famous, and it is famous as something absurd. Strangers are performing impressions of your voice acting, because it is your voice, there was nobody else to do it. The thing you made with your mother is a meme. And at the same time, the money is real, the audience is real, and a substantial number of people in those rooms are crying at the end of a film about a calf.
There is no clean emotional category for that. Vindication and humiliation arriving in the same envelope, on the same day, from the same crowd.
What is also true is that the film reached an audience that no conventional path would ever have given it. A two-person animated feature built in architectural software by a landscape designer and his mother does not get a marketing campaign, does not get festival slots, does not get a critical champion. Under normal conditions it plays to those two hundred and thirty-six people and disappears, and nothing about it changes, and the five years vanish. The ridicule is the only distribution it was ever going to get. That is an uncomfortable thing to write down and it is the honest reading of what happened.
So the ending is the reversal completing and then holding, which is the unusual bit. Novelty audiences usually evaporate within days. This one did not collapse the moment the joke got old. The gross kept climbing for weeks, well past the point where curiosity alone explains it, which means a meaningful number of people were going because other people had told them the film was actually worth sitting through, not because it was funny.
A film that took about a thousand dollars in nine days finished somewhere in the region of nine million. It was written, directed, modelled, animated, edited, scored and voiced by two members of one family in a north-eastern port city, on used computers, in a program designed for drawing buildings, across five years in which nobody was waiting for it.
They were mocked into visibility. Then people watched the film, and some of them stopped laughing, and the tickets kept selling anyway.