The short drama industry has long been plagued by the problems brought by Hongguo.
At noon on February 11, 2026, Li Tao, founder of Xi'an Fenghang Culture, received a call from the liaison lead of a short drama platform, with a very brief message: all projects that have not started shooting will be suspended.
Three years ago, the short drama *Wushuang* produced by Fenghang Culture exceeded 100 million yuan in recharge revenue within 8 days after launch, becoming a landmark event in the industry; by last year, the company's production capacity of live-action short dramas had reached a maximum of over 100 episodes per month, with nearly 500 employees. The Spring Festival schedule that followed was supposed to "stockpile finished dramas" as usual, but right after the call, the entire production line stopped dead in its tracks.
The system that produced more than 100 episodes a month had enormous inertia — with an abrupt brake, figuring out how to halt operations, how to handle refunds, and how to reassure teams were all thorny issues. Li Tao and the senior management spent the entire holiday discussing how to respond after the Spring Festival.
Right after the Spring Festival, on March 3, information related to "Hongguo suspended many projects" became a trending search, and the industry kept spreading news that Hongguo Short Drama (hereinafter referred to as "Hongguo") had paused a large number of live-action short drama projects and revoked the guaranteed payment commitments for production partners.
The so-called "guaranteed payment" was the foundation of this business over the past two years. The platform would cover the production fee at an agreed amount in advance. A short drama usually costs 200,000 to 500,000 yuan, and the highest cost can exceed 1 million yuan. Generally, production companies can earn a 5% to 8% profit from contracted production, while the platform takes the rest to bear market risks. If a drama becomes a hit, the production party can also get revenue share of the project according to the contract. It is rumored that there are "less than 10" production companies nationwide that hold the scale guaranteed quota from Hongguo, and all of them received notifications around this point in time.
The day after this call, on February 12, ByteDance officially released its video generation large model Seedance 2.0, pushing the price of video generation down to a few cents per second.
The trajectory of the industry has changed. Statements like "cost reduction", "sudden huge profits", "produce a drama in 3 days and recoup costs in 5 days", "everyone can make dramas" are circulating on the platforms. With technological upgrades and platform policy adjustments, it seems that the short drama ecosystem is about to usher in a new round of explosion.
Before that, the fastest runner in the AI short drama myth was Huang Haonan, founder of Jiangyou Culture. At its peak, 8 of the top 10 titles on Douyin's comic drama chart were produced by his company; at the end of last year, China Literature approached him with IP resources and investment, pushing the company's valuation to nearly 2 billion yuan. "It's a money printer," Huang Haonan said, adding that the company's revenue in 2025 reached about 800 million yuan. At that time, he could still enjoy the last period of good days, because two months later, he would have to lay off hundreds of employees from his company with a thousand-person workforce.
"At that time, the platform went around telling everyone: do you know that Jiangyou made tens of millions from AI? They promoted this story everywhere, one company after another," Huang Haonan said. "They treat people as 'consumables'." Setting a benchmark to attract a large number of people to enter the industry will result in 80% of them being "buried with the project", and only 20% can make money.
Before the brake was pressed, this industry had grown large enough. As of December 2025, the number of micro-short drama users was close to 700 million. The market size rose from less than 4 billion yuan in 2021 to 500 billion yuan in 2024, and exceeded 1 trillion yuan in 2025.
But on the other hand, the supply side is highly concentrated. According to QuestMobile data, in February 2026, the overall monthly active users of the short drama industry reached 718 million, while the monthly active users of Hongguo Short Drama alone exceeded 300 million. In July this year, Hongguo's daily active users even reached 168 million, a year-on-year increase of 107%, surpassing the sum of daily active users of iQiyi, Tencent Video, Youku and Mango TV.
According to people close to Douyin, from the end of 2025 to 2026, Hongguo has been continuously integrating into Douyin's parent ecosystem: the creation platform has merged with the Douyin Short Drama Copyright Center, with full interconnection of content, contracts and data. Companies in the industrial chain will face a unified set of rules and a single decision-making center. Every turn of the platform is a "visible hand" that pushes people to different destinations.
The impact first hit cities with the largest production capacity. From Xi'an to Zhengzhou, the number of crews starting work per day at each base dropped from more than a dozen at the peak to only one or two. In the first quarter of 2026, the number of live-action short drama projects that started shooting fell by about 75% year-on-year.
Since March, companies that have suspended filming have begun to collectively shift tracks. Leading production companies in the real shooting era such as Xi'an Fenghang Culture, Zhuoyuan Film, and Zhengzhou Sansheng Wanwu have shifted most of their production capacity to AI. In the first 8 months of this year, 430,000 micro-short dramas were launched across the whole network, 13 times the number of the whole of last year, of which AI-generated dramas accounted for more than 90%.
Theoretically, with technological progress and cost reduction, the next step should be an explosion of creativity and profitability. However, according to DataEye data, taking 50 million views as the break-even benchmark, less than 1.3% of AI dramas in the first half of the year could recoup costs, and roughly only 1 out of every 77 new dramas can break even. Seven months later, the most heard phrase in the industry is "no profit". In half a year, the most frequently asked question in the industry has changed from "how to produce faster and cheaper" to "whether to continue doing this and what to make".
Why, after short dramas switched from live-action to AI, with obvious cost reduction and efficiency improvement, has it become harder to make money?
Since its birth, the short drama industry has been a highly stratified industrial chain: the upstream is web novel IP and scripts, the middle is platforms, production companies and distributors, and the downstream is traffic investment and audiences. Over the past three years, the power structure of this chain has undergone two drastic changes: First, the free model turned almost all links into suppliers of the platform, and then AI made the "production" link itself lose its scarcity.
Deeper changes have taken place in the accounting books. The income of every link in the chain is ultimately calculated by the platform, determining which traffic pool a drama enters, at what coefficient it is settled, and how much a single view is worth. The fluctuation of the "settlement coefficient" affects the nerves of all practitioners.
Wang Xiaoquan, co-founder of Zhaotu Culture, has experienced the entire process of this round of ups and downs, and he still insists on focusing on live-action dramas. "Everyone is not doing live-action dramas now, so my work may have scarcity instead," he said, "The threshold for making AI dramas is very low, but the threshold for making good AI dramas is extremely high."
In the past seven months, the industry has verified the first half of the sentence, and the second half has only just begun.
New Rules
It is understood that Seedance 2.0 has pushed the generation price down to about 0.6 yuan per second for 720P resolution, and about 1.5 yuan per second for 1080P resolution. The production cost of AI comic dramas has been reduced to 800 to 1200 yuan per minute, about one-fifth of the cost of live-action short dramas; the cost of an AI realistic human drama has dropped to 60,000 to 100,000 yuan, while a live-action drama of the same scale costs more than 500,000 yuan. For "high-volume production dramas", the above costs can be further compressed several times.
This "cost reduction narrative" has repeatedly become a trending search at the beginning of this year. Companies have continuously exposed their updated workflows: shorter cycles, lower costs, fewer personnel — 3 people, 5 days, one drama is regarded as the new industry standard; every iteration of the large model is accompanied by a lower unit price. In just half a year, the production quotation for standardized AI short dramas has dropped by more than 90%.
This directly detonated production capacity. In June, Douyin alone launched 37,700 new AI dramas, with an average of more than 1,200 episodes per day. However, the expansion of supply and the dilution of returns happened simultaneously: roughly only 1 out of every 77 new dramas can recoup costs. It is understood that in the second quarter, the platform's content review rejection rate once reached 70% to 80%.
But all of these are hindsight. The industry quickly reached a consensus at that time — AI is the next hit. As soon as the Spring Festival was over, everyone was doing the same thing: switching tracks.
Li Tao shifted the company's full focus to AI. This was not easy: nearly 100 employees left one after another; two partners also changed their jobs to serve as the overall leads of AI projects. Without ready-made tools, the team simply developed their own. To lock in computing power resources, Fenghang signed an annual framework agreement: the initial negotiated amount was 10 million yuan, but it rose to 15 million yuan when it was about to be signed, and several groups of agency companies came to compete for his signature. "I can't fall behind," he signed it anyway. Looking back now, there was a possibility of "hunger marketing" tactics at that time.
Others turned around even earlier. Liu Yihong, founder of Xi'an Zhuoyuan Film, studied animation, and his company was formerly a film and advertising company. As early as March last year, he led seven or eight people to test AI comic dramas. His first work *Myriad Worlds Bank* received more than 20 million views and recouped costs. He expanded his team, and the AI team grew to 50 or 60 people by July, capable of producing 20 to 30 episodes per month.
In July last year, Wang Hongfang, founder of Zhengzhou Sansheng Wanwu, also selected more than ten people from his 100-person post-production team to try to make comic dramas with AI. There were no courses to take and no one to teach them, so they tested every tool on the market one by one, polishing the prompts word by word.
In the first six months, they "hardly made any money", but the direction was set: they would no longer focus on contracted production, but shift to self-owned copyrights. He soon drew his own conclusion about this wave: AI can only make money for people who truly understand it.
Source: AI Generated
As for what "understanding" means, Wang Hongfang believes that operation, resources and technology are all indispensable. Those who only understand technology cannot make money, and those who only have resources are not enough. In the end, what is tested is operation, that is, the ability of distribution and monetization — "Where you put it, it can turn into money."
"Many people regard the capabilities of the model as the capabilities of their own company, which is a common problem among small companies," he said. Technological equality only levels the playing field of tools, but not the gap of cognition. "Only people with cognition can make money, and those without cognition are all 'cannon fodder'."
But the number of entrants is still increasing. A local business owner came to consult Wang Hongfang, wanting to enter the AI short drama industry, considering it a trend. He had made a lot of money before, and now he wanted to "go all out" — renting an office in a prime location and recruiting nearly 100 people to make AI short dramas. Wang Hongfang poured cold water on him face to face, and his suggestion was always "small team, small scale". He even calculated the accounts for the other party: target a team of 20 people, operate at a loss for 6 months, "only lose 300,000 to 400,000 yuan, and figure out how this business works first". But the other party did not listen, and later lost more than 2 million yuan in only 3 months.
The production organization is being restructured. In the live-action drama era, a crew is a team of about 40 people with extremely detailed division of labor; an AI drama crew shrinks to less than 10 people, or even 2 to 3 people. The focus of the workflow is also shifting. The production of AI dramas is actually derived from the "post-production" end: the editor first figures out what images are needed for each segment, then generates materials, and fills in the gaps as needed.
In Wang Hongfang's view, there are only two viable ways to survive in this round: either top companies that are extremely "understanding", or super individuals that are extremely small — "a company with 3 to 5 people has a way out". Technological equality puts tools in everyone's hands. "Ordinary people can make IP-based content with low cost and polish it slowly", while companies in the middle that rely on multi-layer subcontracting for contracted production are losing their positions.
Some people choose to stay on the live-action drama side. In fact, Zhaotu responded very early: in the spring of 2025, Wang Xiaoquan cooperated with a game IP to make comic dramas, but it did not succeed; in the summer, he expanded his AI team to 70 or 80 people. But he could not make up his mind whether to fully transform to AI. Zhaotu is an asset-heavy company, and a large number of assets accumulated over the years of live-action shooting are like a burden on its back; a team of hundreds of people cannot be disbanded easily.
At present, Zhaotu has shrunk to more than 100 people, producing 20 to 30 live-action dramas and 5 to 6 AI dramas per month. It has transformed from a top supplier of Hongguo to a smaller, more flexible company. "Don't act blindly before you find a clear direction," he said, the priority is to stay alive.
Changes at the Most Upstream End
The scripts of short dramas are highly dependent on web novels.
Industry data agencies estimate that more than half of the short dramas are adapted from web novel IP, and the proportion in hit works is as high as 80%. The adaptation licensing fee for S-level web novel IP ranges from 80,000 to 100,000 yuan, popular A-level works cost 60,000 to 80,000 yuan, and the most ordinary ones cost more than 30,000 yuan. China Literature, Fanqie Novel, ChineseAll Digital Publishing and Migu Digital Media have all turned their IP libraries into "script pools" for short dramas.
In the past two years, in order to lock high-quality scripts into its own library, Hongguo raised the remuneration standard for screenwriters: the guaranteed payment for top screenwriters once reached 150,000 to 200,000 yuan, with additional revenue share. In contrast, other platforms can hardly offer more than 120,000 yuan as guaranteed payment. The platform's mechanism gives money to both ends: screenwriters upload scripts to the library, and they can get paid only if their scripts are selected; production partners have the right to choose scripts and receive production fees.
This mechanism quickly created arbitrage opportunities: some people uploaded "connection-backed scripts" to the library, had their own people select them, and the money would circulate once the contract was signed. Many practitioners mentioned that the platform had differentiated policies for different companies, and a few large companies were once almost "exempted from inspection", but they were unwilling to name any company. They only said that there was too much money in 2025, and everyone was taking advantage of the loopholes, "the platform closed the loopholes after settling accounts at the end of the year".
AI has made the definition of copyright more complicated. A large number of AI realistic human dramas use faces similar to celebrities, leading to a sharp drop in review pass rate; "AI manuscript washing" that retains the core framework of hit dramas and remakes them with modified details