AI short dramas are experiencing a carnival where "the more popular they become, the harder it is to make profits"
Recently, film and television stocks seem to have regained their long-lost excitement.
On August 31, *The New Journey to the West*, the first AIGC-produced long drama in China to be broadcast on the prime time slot of a satellite TV, premiered on Mango TV, and landed on Hunan TV's prime time the same night. The launch of this drama directly ignited the capital market: Mango Super Media's stock price hit the 20% daily limit for two consecutive trading days on August 31 and September 1, with its market value surging by more than 10 billion yuan in just two days.
Soon, capital began to look for the next investment target along the track of "AI short drama".
Companies including Huanrui Century, Dasheng Culture, Zhongguang Tianze, Bona Pictures, Yaowang Technology, and Zhangyue Technology saw their share prices rise one after another. Among them, Huanrui Century recorded 4 daily limit gains in the 5 trading days from August 31 to September 4, with a cumulative increase of 42.97%.
The Hong Kong stock market has seen even more drastic movements.
The share price of Duiba surged by up to 336% in 9 trading days.
If you only look at the K-line, this looks very much like a new round of rally for film and television stocks, but after checking the semi-annual reports of these companies, you will find that the situation is not that simple.
On the evening of September 4, Huanrui Century issued a stock trading risk warning announcement, clearly stating that the company's AI short drama business has not yet formed a scale effect in terms of the number of launched titles, revenue scale and cost-benefit conversion. More importantly, in the first half of 2026, the revenue from AI short dramas accounted for less than 1% of the company's total operating revenue.
For the same AI short drama track, Duiba has delivered a completely different performance.
The company launched its AI short drama business in January this year, and the relevant revenue in the first half of the year reached 223 million yuan, accounting for 51.2% of its total revenue. In the same period, the company's total revenue was 435 million yuan, up 24.34% year on year; its gross profit was 74.54 million yuan, up 30% year on year; its net loss was 25.33 million yuan, narrowing by 5.2% year on year.
Obviously, the capital market has already priced in the positive prospects of AI short dramas in advance, but there are not as many companies that can actually translate this growth story into their financial reports as people might expect.
01
The Real Threshold of AI Short Dramas Is Not AI Technology
On the evening of August 31, the first season *Huaguo Mountain Chapter* of *The New Journey to the West* was officially launched on Mango TV, and landed on Hunan TV's prime time slot.
For the first time, AI-generated content entered the most mature content consumption scenario of traditional TV stations in the form of a relatively complete long drama.
In the past two years, AI-generated videos are no longer a novelty. A large number of AI short videos are shared on social platforms every day, with each shot only a few seconds long, and batches of materials can be generated in just a few hours.
But there has always been a threshold between these fragmented materials and a complete TV drama.
A TV drama cannot be made by simply piecing together several nice shots. The characters need to maintain consistent images, the scenes need to be coherent, the audio needs to be unified, the plot needs to move forward logically, and the dozens of minutes of content also need to keep audiences willing to continue watching.
Therefore, the release of *The New Journey to the West* is more like a concentrated demonstration of industrial capabilities. AI has begun to evolve from "generating a single shot" to "producing a complete work", and short dramas happen to provide the most suitable test field for this capability.
Traditional long dramas usually have dozens of episodes, with long production cycles and huge investment in a single project.
Short dramas are completely different: themes can be tested quickly, scripts can be adjusted rapidly, and user feedback can be obtained in a short time.
If the performance data of the first episode of a drama is poor, the subsequent content can even be adjusted in time.
This production mode is particularly important for AI. Because the biggest advantage of AI has never been "producing a perfect work", but being able to generate many versions at low cost and high speed.
In the past, when a film and television company shot a drama, it had to confirm the actors, directors, scenes and budget first. Once the shooting started, the cost of modification would be very high.
Now it may become a different mode: make a first version to see if users like it, then adjust the characters, plot and even the ending, and continue subsequent production.
This is somewhat similar to the iteration logic of Internet products.
More importantly, short dramas already have mature business models, including user paid unlocking, advertising, e-commerce, platform revenue sharing and overseas distribution. These models are not created by AI, AI only makes the original production links faster and cheaper.
Therefore, it is not surprising that AI short dramas are easier to generate revenue than many so-called "AI applications".
This point can be clearly seen from the performance of several companies in the first half of this year.
Duiba is the most direct sample. This company used to focus on Internet marketing and SaaS business, but AI short drama has quickly become an important source of its revenue.
The data of Kunlun Wanwei is also impressive. The company's short drama business revenue in the first half of 2026 reached 1.534 billion yuan, up 163.28% year on year; by the end of June, its monthly gross flow exceeded 65 million US dollars. More than 90% of the newly launched content on its DramaWave platform in July was generated by AI.
China Literature has taken another path. In the first half of 2026, the company's revenue from short dramas and AI manhua dramas reached 430 million yuan, up 230% year on year; it has adapted thousands of online novels into AI manhua dramas with the help of AI, among which 46 works have exceeded 100 million views and 367 works have exceeded 10 million views.
On the surface, these companies are all working on AI short dramas, but in fact, the resources they rely on are completely different.
Duiba is more like using AI to enter a new content track; Kunlun Wanwei relies on its overseas platforms and distribution capabilities; behind China Literature are millions of online literature IPs.
Putting these three cases together illustrates a point: AI is only a production tool, and other assets held by the company still determine its real commercial value.
02
AI Has Reduced the Production Cost of Dramas, and Made Competition More Fierce
The biggest advantage of AI short dramas is that they reduce production costs.
According to the research report of Zhongtai Securities, in the first half of 2026, the production cost of 2D/3D manhua dramas and simulation real-person AI short dramas dropped by 80%~90% on the whole compared with 2025, which is only 2%~10% of the production cost of traditional short dramas.
However, the decline in cost has not brought profit release, but accelerated the influx of participants on the supply side.
In the first half of 2026, the number of new native AI dramas and manhua dramas launched on Douyin reached 221,900.
According to the forecast of Zhongtai Securities, the domestic output of AI short dramas will increase significantly from 60,000 in 2025 to 1 million in 2026. The exponential expansion of the supply side means that the probability of each new drama being noticed by users will be even lower.
The film and television industry often had a saying in the past: "Content is king"; in the era of short dramas, there should be an additional sentence behind this saying: "Content is king, but you have to be noticed first".
An ordinary AI short drama, even if its production cost is only one fifth of that of a real-person short drama, still requires investment in traffic promotion, distribution and user conversion.
If no one clicks to watch it, the low cost is still a waste; if someone clicks in but exits after three minutes, the high generation efficiency makes no sense at all.
Therefore, a subtle change is taking place in the AI short drama industry: the production cost on the set has decreased, but the traffic cost may not decrease, and may even become more and more expensive. That is why in the future, the AI short drama industry is likely to follow the rule of "the one who is best at picking out hit works from a large number of content wins" rather than "the one who produces the most works wins".
This is very different from the traditional film and television industry.
In the past, a film and television company that produced ten dramas a year could afford the failure of one or two works; in the AI era, a company can totally produce hundreds of dramas a year.
Higher failure rate is not scary, the key is whether the remaining few successful works can generate enough profits.
As a result, AI short drama operation is increasingly like a "portfolio investment": produce 100 dramas, 90 of which perform averagely, 9 of which barely recover the cost, and 1 of which becomes a hit.
This is also why data capabilities and distribution capabilities are becoming more and more important. You need to know what themes are gaining popularity recently, what character settings and plots are easy to retain users, and which work is worth continuing to invest in traffic promotion.
AI models can help analyze these data, but these rules will not be created out of thin air by the model, they require a large amount of real user data for training.
In the future, the truly valuable companies may not be the ones that are "best at generating videos", but the ones that know their users best.
This is why companies like China Literature are worth special attention, because they own a large number of online literature IPs. In the past, it might take a long time for a novel to be adapted into a film or television work, but now AI can quickly turn it into a short drama.
Therefore, if we look further ahead at this round of AI short drama market rally, the most noteworthy point is how much real revenue AI short dramas can bring to a listed company, which may be the dividing line for the entire industry to shift from "concept-driven rally" to "performance-driven rally".
Huanrui Century has given a very clear answer: the revenue is not large for now.
Mango Super Media is in a different situation. The company has established an AIGC innovative content center and an AI industrialization center, and continues to increase the application of AI in content production. In the first half of 2026, the "Big Mango Project" launched a total of 2647 micro short dramas, up 124.51% year on year; the average daily DAU of the short drama channel increased by 104% year on year, but the overall performance of the company is still under pressure.
This shows that even if a traditional film and television platform really uses AI, we cannot simply assume that "profits will come as soon as AI is introduced". Because the final profit statement of a company still has to face a series of practical problems such as member operation, advertising business, content cost and platform competition, and AI can only solve part of these problems.
03
AI Short Dramas Bid Farewell to Unregulated Rapid Growth
Since September 1, the *Administrative Measures for the Development of Micro Short Dramas* has been officially implemented. For micro short dramas generated and produced with artificial intelligence, the regulatory authority requires that AI-generated content labels be added in accordance with the regulations. This means that AI short dramas are evolving from the earliest "unregulated rapid growth" stage to a more mature stage.
In the future, the industry competition will no longer only focus on generation speed or how many works can be launched in a single day, but on who owns high-quality IPs, who has deep user insights, who has the ability to obtain low-cost traffic, who can continuously create hit works, and who can finally turn these advantages into actual cash flow.
Therefore, if we break down this recent market rally, we will find that the market has already completed the first step of development.
*The New Journey to the West* has answered the first question: can AI produce long dramas? The answer is yes.
The semi-annual reports of companies including Duiba, Kunlun Wanwei and China Literature have answered the second question: can AI short dramas generate revenue? The answer is also yes.
But the third question that determines the height of the next round of market rally is: can AI short dramas generate sustainable profits?
The answer to this question will be revealed in the upcoming third quarterly reports and annual reports.
This article is from the WeChat official account "Alpha Workshop Research Institute", written by Su Di, and published with authorization from 36Kr.