The production capacity of live-action short dramas has begun to “collapse” nationwide.
Author丨Reporter Chen Hao of China Entrepreneur
Editor丨Zhang Hao Ma Jiying
Source丨China Entrepreneur Magazine
On the afternoon of August 24, when China Entrepreneur met Shi Yiyuan in an old office building in Zhengzhou, he had just hung up a phone call.
The old air conditioner with a fifth-level energy consumption rating was laboriously running, and the walls and cabinets were filled with team launch commemorative photos and trophies. On the table lay a male-oriented short drama script that had not yet been filmed. This was given to him by an old friend—some platform short drama businesses they frequently collaborated with were suspended for the past two years. His old friend had filmed four pieces himself, all rumored to be hits, and now wanted to pull together a team and invite Shi Yiyuan to invest in production.
The old friend offered over 300,000 yuan and asked him to invest another 100,000 yuan.
“So I was just on the phone asking about the market.” He said, “I haven’t filmed anything since the new year, I have only been working on AI. Now I know everything about the AI market, but if I want to shoot short dramas, I definitely need to inquire from various sources.”
The numbers he inquired returned like this: last year the lead actors earned 8,000 to 10,000 yuan per day, now some are shooting for only three to five thousand yuan; a middle-aged actor he frequently cooperates with reported an external rate of 1,500 yuan last year, and now is accepting offers as low as 500 yuan, even so, he could only work a maximum of 10-odd days a month; previously the daily rate for stagehands was normally 400 yuan, but now it is heard that some are willing to work for 200 yuan a day. The price reduction across most job types has nearly “halved” from the start.
“There’s no other way.” He said, “The big environment is not good, everyone has no work. If you don’t lower your prices, no one will look for you, but even if you lower them, sometimes no one will look for you, which is a waste of effort.”
He counted some acquaintances’ directions on his fingers: an actor who was still starring at the beginning of the year has gone to sell houses; a middle-aged actor who contacted him in April to work in AI is now a security guard; and other friends have opened clothing stores, live streaming rooms, or become Feng Shui consultants…
In the past three years, Zhengzhou has been referred to in the industry as a “short drama hub.” According to industry data, over a thousand micro short drama companies and nearly 40,000 practitioners gathered here, with a market size of nearly 10 billion yuan and an average of nearly 100 productions starting each day. The majority of these are small and medium-sized companies.
In the week leading up to the Spring Festival of 2026, two matters occurred in quick succession: Hongguo Short Drama (hereinafter referred to as “Hongguo”) retracted its guaranteed bottom-line policies, and production companies with large guaranteed quotas nationwide received notifications on the same day—any projects that had not started would be halted; following this, ByteDance officially released the video generation model Seedance 2.0, further lowering the production threshold for AI short dramas. After the Spring Festival, the production capacity of live-action short dramas began to “collapse” across the country.
According to the rhythms of previous years, mid-March should have been the busy season for resuming work. However, Shi Yiyuan has only produced four live-action dramas this year, two of which have not yet been launched. His company’s team of around ten people has taken more than a month of long vacation. “I consider myself a foreman,” he defines himself, “if there’s no upstream work, the downstream can only close down.”
Undercurrent
This “foreman” is 30 years old, from Weishi, Henan, and graduated in 2019 from the Dalian Art Academy with a major in film and television performance. He was born in a rural area and was raised by his grandparents; his parents went to work in other provinces at a young age. Before graduation, he found a job as a short video director in Beijing, with the contract finalized, but he chose to give it up because his girlfriend was in Zhengzhou and his family urged him to settle down.
After returning to Zhengzhou, he initially had no work. He tried applying for a reserve cadre position at the supermarket where his girlfriend worked, passed the interview, and on the day he was supposed to sign the contract, an advertising company called him, inviting him to audition for an acting role. Since he hadn’t signed the contract yet, he went for the audition at that advertising company and was excited to return to his field.
Zhengzhou is a hub for information flow advertising in China. He didn’t understand this at the time, thinking he had joined a “pyramid scheme.” Later, working in it, he found it quite simple; primarily shooting promotional ads for educational apps and various consumer loan products, “I could shoot dozens of them in a day.” The last line of dialogue always was: “Click the link below the video to download.”
Looking back, Shi Yiyuan recalls that he once filmed content that was almost in the form of a short drama—turning climax segments of online novels into videos to guide users to reading platforms. But those videos could not be considered dramas; they were just “hooks”—a product born from the advertising gene designed for conversion instead of craftsmanship from day one. He didn’t truly realize this until years later.
The life of information flow advertisement actors was quite “easy”: a base salary of 4,000 yuan, nine to five hours, weekends off, and when there was no work, they could just rest. He couldn’t suppress the urge within him. When the company was about to convert his employment to regular full-time status with benefits, he resigned. During this time, he took an entrance exam for the directing department of the Beijing Film Academy—without studying at all, joking with the proctor upon submitting the paper: “Teacher, wait for me, I will definitely pass next year.”
Upon returning from the exam in Beijing, on November 12, 2019, he registered Henan Yiyingyuan Film and Television Media Co., Ltd. in Zhengzhou, earning over 10,000 yuan in total by the end of the year, “which was less than a job, but there was a certain adventurous spirit at that time.”
This adventurous spirit helped him endure the COVID-19 pandemic. He not only filmed medium-length videos to earn views and subsidies but also went to the gold wholesale market every day to film “gold price explorations,” earning a couple hundred yuan for helping fans purchase gold. In 2022, as information flow advertising exploded, he could rush to several sets in a day, seamlessly transitioning between “loser characters” and “dominant CEOs,” and at that time, daily earnings could reach 800 or even 1,000 yuan.
At that time, a form of live broadcast drama called “live drama” also began to rise on platforms like Kuaishou: family grudges reached their emotional peak, and the host would start selling goods. He usually played the “villain”: “I had a very deep impression the first time I acted; the host was in a smelly pigsty, and my scene was to kick the door and shout ‘pay back the money’; that day the live stream's popularity was shocking.” These supporting roles could earn him five to six hundred yuan per scene.
In mid-2022, a friend asked him to act in a short drama. Filming a 30-episode series in one day amazed him. There were two cameras focused on two groups of actors: one group acted in front of the camera, while the other group stood nearby learning their lines, and there was someone off-camera providing dialogue cues. Those who forgot their lines had to step down and memorize again, while those who had just learned would act on top, basically all done in one take.
That afternoon at 5 o'clock, he had to rush to act in a live broadcast drama, worried he wouldn’t finish in time, but before it was time, the director let him go. He was very happy, earning money from two jobs in a day, feeling like he struck gold, but didn’t realize he was stepping into an expanding new business model over the next years.
The Last Link
After entering the industry, Shi Yiyuan became a director within a few months.
In December 2022, someone asked him to direct a 100-episode drama, which was completed within five days, with a director's fee of 1,500 yuan per day. On the first day, filming went until three or four in the morning, and early the next morning, he was back to work at six or seven; while biting into a bun that day, he realized it was covered in blood—after staying up all night, his lips had cracked and bled. He heard that the drama invested several tens of thousands of yuan, but the specific investment amount and revenue had nothing to do with him.
This represents the most typical survival form in the short drama industry chain. Most companies in Zhengzhou are involved in short drama contracting, taking on orders from platforms or large companies, earning a price difference. The industry recognizes that pure contracting profit margins range from 5% to 8%, while crews are organizations that burn money by the day—he calculated that once filming starts, the hard costs are at least one yuan per second, and a single rain, an overdue shoot, or the lead actors exceeding time can eat up that profit. The so-called production fee is essentially not profit but the basic price for renting out the team’s time.
Later, Shi Yiyuan figured out that first-level suppliers are not truly selling production capabilities but rather the qualification to receive orders from platforms. The pressure of funding, labor risks, and uncertainties are layered downwards, ultimately falling onto the least negotiable end of the industry chain—only there do real people, real days, and real cash need to be used to fulfill the profits on contracts.
In his first two years in the industry, he also encountered a shift in business models. Initially, short dramas operated predominantly on paid apps, where 80-90% of revenues were used for marketing, with recharge payments also having deductions from WeChat withdrawals. He heard that if the ROI for marketing was around 1.2, one could basically make money, while the deduction rates were constantly rising, consuming most of the remaining profits.
Later, Douyin cut off the channel to WeChat mini-programs, retaining that traffic for its own ecosystem. In August 2023, Hongguo launched and ran with a free model for a year, and by the second half of 2024, free and paid dramas accounted for half the market, turning past the new year, free dramas became mainstream. Shi Yiyuan had no concept of Hongguo’s size; he only knew most of the dramas he launched ultimately ended up in the libraries of such free platforms.
In 2023, he joined a newly established company as an executive director, with a salary structure of over 10,000 yuan base plus a few percent in profit sharing, which he valued. After joining, as long as he was shooting in the team, he needed to work 16 hours a day, averaging only three to five hours of sleep, suffering from palpitations and shortness of breath three times. The immense pressure was difficult to bear, and after three months, he resigned.
Later, just as Migu wanted to shoot short dramas, they were the first to find him. The first piece was a period drama, adopting a shared cost and shared income model, with Migu covering most of the costs. However, there was a rule in the Hengdian filming crew that required pre-payment before filming; due to time constraints, the filming was cut from seven to six days, but he squeezed it in while ensuring quality. The second drama, Junlin Tiantang, took a full seven days to shoot, but when it went online, the data did not perform well, and he didn’t earn any profit share, “just the director's fee.”
The third drama was Gai Shi Yi Zun, which was the closest he had come to money. The second round of airing for this drama was on the Hongguo platform, achieving results, and he estimated that the copyright share paid to the rights holders could be over a million yuan. But that money had nothing to do with him—he only agreed on the first round of profit-sharing rights, and due to disputes with the partnership company, he also did not receive the director's fee, with the only thing he got being a trophy for the Annual Director Award from Migu.
From the perspective of small and medium contracting companies, platforms can roughly be divided into two types: Migu represents one type, where each project requires a bid-winning process, which is relatively standardized, meaning the profits are divided among many people, strictly controlled to the extent that a month can only allocate one drama.
Hongguo condenses quotas for a few leading companies, which cannot finish all the work they have, accepting hundreds of orders from the platform each month. Each crew can shoot up to three dramas per month at maximum workload—production capacity cannot support the orders, and unfinished orders can only be redistributed downwards. Each layer of subcontracting means prices are “peeled” off layer by layer until they fall into the hands of those with no room for subcontracting.
In 2025, Shi Yiyuan began taking on work from a large contracting company. That company could produce over 100 Hongguo dramas a month, and he could only get one or two of them; he clearly understood his position in the chain.
He compared Hongguo’s subsidies to the “ride-hailing wars” of the past: burning money to grab market share, and a retreat was inevitable. For companies below the mid-tier, payment terms always dangled like a sword over their heads. By the end of 2025, when negotiating cooperation with a leading production company, the payment terms had lengthened, and when settling at the end of the year, he heard that the other party even borrowed money to pay the final bill. He remembered the relationship but dared not proceed—prices were already low, and extending payment terms elevated the risk to the point where it was not worth the gamble.
Deliberation
After the Spring Festival, Shi Yiyuan began working in AI. There were two reasons—live-action dramas had not started; two friends in AI invited him to partner. “I wasn’t really interested in AI, nor was it my initiative, but since there was no work for live-action short dramas, it was somewhat forced upon me,” he said.
He first investigated several of his friends' companies and spent money on tutorials—by March he formed a team, in April he refined the process, and he even worked through the May Day holiday, as he had officially received his first job. He wasn’t very early; a brother company downstairs had already gone to Guangzhou to study in the previous year, “last year was for learning, this year was for proper execution, but I heard they haven’t made much money either.”
The job was a 2D comic drama produced by an online literature platform, totaling 180 minutes, with a contract rate of about 400 yuan per minute. In practice, the first six episodes, which were about 12 minutes total, consumed 10,000 yuan in computing power, equivalent to nearly 800 yuan per minute, nearly double the contract price.
Money burned in card drawing. AI-generated images needed to be consistent—if a character wore glasses in one shot, the next shot couldn’t have them gone, watches could disappear, and zippers might unfasten; if there was a mistake, it would need to be regenerated. By the time the whole piece was completed, computing costs amounted to over 60,000 yuan, not including music, sound effects, and dubbing, while the total budget was only over 70,000 yuan.
“In live-action dramas, modifying footage costs manpower and time. In AI, each generation costs money; every revision requires payment, and clients are sometimes indefinite in their modifications.” Shi Yiyuan believes this is the most essential difference between AI dramas and live-action dramas, as the marginal cost of AI dramas is rigid. However, clients do not perceive this distinction; they only understand AI is quick, thus making more modification requests than before. AI indeed increased output speed while also lowering the difficulty of clients deciding to “scrap the previous version.”
The second job required a horizontal-screen replica drama that needed to be delivered in 1080P—every increase in resolution raised both the price and the waste film rate. The other party offered over 1,000 yuan per minute, but he dared not accept it: having learned from the previous job, he was afraid the computing power would exceed the budget and proposed that the computing power be separately calculated, with himself charging production fees of only four to five hundred yuan per minute.
Five people worked for a week, burning nearly 30,000 yuan in computing costs, producing over twenty minutes of footage, which was completely rejected by the client, leaving only three minutes. Later, he simply changed to fully charge based on person/day, at 400 yuan per person per day, regardless of whether the minute was usable or not. Ultimately, after over half a month of work, he earned over 10,000 yuan. Looking back, this was the best decision he made this year.
What made him even more uneasy was not the money. In the live-action era, at least he knew if he was in the “three-pack” or “four-pack,” while in the AI era, he didn’t even know that, “Who are we working for? Whose money are we earning?”
Subcontracting in live-action dramas is “clumsy,” requiring venues, crews, and a group of known individuals, making the chain visible. In contrast, AI subcontracting requires none of this; a single mass message or online video can complete a transfer regardless of acquaintances—thus the chain has become efficient yet invisible. He transitioned from being proportionally deducted to being silently “diluted.”
At the same time, the industry threshold has also collapsed. He has seen couples from Sanmenxia, who went to Guangzhou for a few days learning AI, returning wanting to start a company; senior peers involved in car beauty also asked him if they could work in this area; wedding companies, engineering firms, and bosses in new energy flocked into the field... In his words, anyone with any connection to video wants to enter the industry, “and at this stage, those truly making stable profits are those selling tutorials and conducting training.”
“Those who haven’t entered can’t see the pitfalls, while those who have entered have already stepped in.”
He has also scrolled through industry data from the first half of the year: only on Douyin, 221,900 new AI short dramas were added, with only 1,055 hitting over 100 million views, less than 0.5% in the billion-view rate. In the face of such a denominator, the seven years of experience he accumulated began to lose efficacy.
The platform’s attitude towards AI content has also tightened. On July 22, Hongguo released an announcement regarding the “high frequency of AI faces,” image homogenization, and copyright infringement, launching a special rectification. Over half a year, AI dramas transitioned from being popular to being regulated.
Some peers have also advised him to create self-produced dramas and overseas dramas to try for a hit; he shook his head, “If self-produced dramas fail, the losses will be greater, and if everyone makes overseas dramas, the money won’t be made anymore.” Those who can stabilize their businesses either have stable quotas from the platform or their own distribution channels for marketing, and he has neither of those.
The interview concluded that evening, and a group of people went out to eat. At the dinner table, Shi Yiyuan’s WeChat chimed. Someone found his contact through connections and asked if he could bring them into making AI short dramas; he directly dissuaded them. It is not that he is conservative or unwilling to bring people along, but he himself does not know where the money is. Seven years ago, he sought jobs everywhere; now when someone asks him for guidance, he feels that the most valuable thing he could offer is advice.
He is 30 years old, with two children, the elder soon going to primary school, his mother in Zhengzhou helping him with the kids, and his father still working in another province. He is not as “brave” as he was back then.
When asked what he plans to do next, will he enter the live-action drama scene again? He thought for a long time, first describing the current situation he observed: live-action short dramas are actually slowly warming up again, with some starting production, just not many.
“This industry has淘汰了 many people.” He spoke in a calm tone, “I don’t know if I can wait for the tide to come back. If there is work, I will take it, but I have to consider carefully first.”
This article only represents the author's views and does not reflect the position of 投中网
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
