China's first AI theatrical film opens in theaters, grossing just over 300,000 yuan on its first day
On October 23, China's first AI theatrical film to receive the Dragon Seal officially opened in theaters, grossing just over 300,000 yuan on its first day, with Lighthouse projecting a total box office of 2.169 million yuan. The producer said the cost was in the tens of millions of yuan, and that under normal revenue sharing the break-even line is about 25 million yuan; Bona Film Group CEO Jiang Defu responded that it was not generated with one click, and that a team of nearly 100 people spent more than two years doing frame-by-frame fixes according to the film industry workflow, discarding 1.2 million base images. The article also mentions iQIYI's Liaoyuan Plan, ByteDance's Jimeng Studio, and other AI feature film initiatives at home and abroad.
Why read this
AI score 56
From the opening-day box office, the tens-of-millions-yuan cost, and the producer's claim of frame-by-frame rework, one can see the real market and production gap once AI feature films enter theaters.
Topics
Body · AI translation
Lost a fortune? The first domestic AI theatrical film hits theaters
Tech Fox October 25, 2026 08:43:52
On October 23, the first AI theatrical film in China to receive the Dragon Seal officially premiered.
The premise is quite interesting: the mysterious carved symbols on Sanxingdui bronzes actually conceal an ancient binary code.
Put plainly, it takes the Sanxingdui IP, which carries an air of mystery, and wraps it in a sci-fi shell.
Audiences widely complained that the characters' facial expressions were stiff and their movements unnatural, and that watching for a while produced an indescribable sense of incongruity.
Some also joked: the only people buying tickets are probably film and TV bloggers.
So the box office, naturally, was less than ideal: the first-day gross was just over 300,000 yuan, and Beacon predicts a total of 2.169 million yuan.
Many assumed the budget must have been especially low, but it may be higher than people imagine; the producer says it is in the tens of millions of yuan.
With a production cost in the tens of millions, under normal box-office revenue sharing the break-even line is roughly 25 million yuan.
Given the current box-office trajectory, losing money is basically a foregone conclusion; after all, not everyone has the good fortune of Niu Lai.
The production side is also feeling aggrieved, and Bona Film Group's chief administrative officer Jiang Defu came out to plead their case.
He said not to believe in any one-click generation: the team numbered nearly a hundred people, followed the film-industry workflow to retouch frame by frame, worked for more than two years, discarded 1.2 million base images, and scrapped and redid the entire film over a hundred times.
To court theaters, they pushed the distribution price down to 20 yuan and conceded the revenue split to 40-60, whereas normal domestic films are 43-57, effectively giving theaters three extra points.
Although they took a very humble stance, the box office still would not budge.
Because at the end of February this year, Tuan Yuan Ling had already been released, an animated film based on the giant pandas Tuan Tuan and Yuan Yuan, billed as China's first full-process AIGC animation.
But that one skewed toward younger audiences, drew less attention, and did not spark such a large industry discussion.
Bona set up "Boka Cloud Studio"; Sanxingdui was only the first trial, and Plan Extra: Guanghan Operation and Chinese Strange Tales: Luocha Sea Market are already lined up.
In this respect, Mango TV has consistently been ahead. At the end of August this year it launched Hou Xi You Ji, the first AI long-form drama in Chinese television history to air on satellite TV.
Its premiere ratings ranked first among provincial satellite channels in the same time slot, its online reach approached 3 billion, the main feature drew 150 million views on Mango TV, and Mango Super Media's stock hit the daily limit for two consecutive days.
Bear in mind that the cost was only 900,000 yuan per episode, less than one-tenth that of a live-action drama in the same slot.
iQIYI launched the "Liaoyuan Plan," aiming to release more than forty AI films this year, and built a dedicated production platform called Nadou Pro that covers the entire process from screenwriting to editing.
ByteDance also launched "Jimeng Studio" at the end of August, dedicated to incubating AI feature films of ninety minutes or more; S-level projects get their full production costs covered, along with distribution and film-festival resources, making clear its intent to spend money to get this path working.
Don't think only China is racing on this; overseas is not idle either.
There is also an AI sci-fi blockbuster called Raphael, set for release on November 25, built around a future-war worldview.
In North America, Gods Don't Give Gifts is set for release on December 4, and it has explicitly said it will submit for the Academy Award for Best Animated Feature.
Put plainly, this is not one company's gamble; the entire global film and television industry is feeling its way across the river.
This year the entertainment world has been having babies left and right; since there is nothing to shoot anyway, idle is idle, and netizens have been joking that the domestic entertainment industry has entered a "Great Pregnancy Era."
Fox Sister still remembers that in March this year, "AI actors for roles below second male lead" trended on social media, and at the time there was an outcry.
Even a short-drama actress who had appeared in a hit with over 100 million views went three months without landing a single role, and some male actors simply started setting up street stalls.
According to industry insiders' predictions, in the next five to ten years at least thirty to fifty percent of professional positions in the film and television industry will disappear. It is not a temporary lack of work; it is structural disappearance.
In fact, what disappears is not just the actor's job, but the entire ecosystem that supports those jobs.
It is just that we are still at a very early stage, much like when silent films first appeared a hundred years ago: crude technology, immature storytelling, and audiences still adapting.
That much has not changed in the past hundred years, and in another hundred years, it most likely will not change either.