In the era of AI, long - form video platforms are struggling on the verge of survival.
In 2025, it was a terrible year for long - form video platforms. To be precise, every year since 2020 has been tough for them, but last year was particularly bad:
iQiyi's annual operating revenue declined by 7%, and its net profit after excluding non - recurring items dropped by nearly three - quarters. The annual revenues of its three major businesses, namely membership, advertising, and content distribution, all showed a downward trend.
The number of paid members of Tencent Video stopped growing and lagged behind Tencent Music. Although Tencent Video achieved profitability, like iQiyi, it was achieved through cost reduction and efficiency improvement.
Mango TV, as the "real third - largest long - form video platform" (with a higher MAU than Youku), saw a decline in revenue in the first three quarters. The pre - announced net profit after excluding non - recurring items for the whole year is expected to decline by 33 - 47%.
Interestingly, if we consider Bilibili as a "long - form video platform" (note: Bilibili can't wait to get rid of this label), it was actually the best - performing one in this industry in 2025. However, even for Bilibili, the user watch time only increased slightly in 2025, and the main driving force for its turnaround from losses to profits was the game business. From 2018 to 2022, Bilibili once invested heavily in traditional long - form video content, but it has pulled back in recent years, which is one of the reasons for its turnaround.
The collective struggle of long - form video platforms is almost synchronous with the decline of the domestic long - form drama industry. As we all know, the lifeblood of long - form video platforms is dramas, followed by variety shows. As for movies, live events, etc., they are just embellishments. In terms of attracting paid subscriptions and providing user watch time, the position of dramas is irreplaceable. The domestic long - form drama industry has been in a long - term decline in recent years, and the situation in 2025 was particularly bleak:
According to the statistics of Yunhe Data, the effective playback volume of the top 20 newly launched long - form dramas in 2025 was 29.6 billion, a 20% reduction compared to 2024.
The drama with the highest playback volume among the newly launched long - form dramas in 2025 was "The Legend of Hidden Sea", with only 2.23 billion views, more than half lower than the first - ranked drama in 2024, "Joy of Life 2" (the latter had 4.51 billion views).
In 2025, no newly launched long - form drama had an average single - episode playback volume exceeding 100 million, and even those exceeding 50 million were rare, breaking the trend of previous years (the last drama with a single - episode playback volume exceeding 100 million was "Joy of Life 2").
Even more terrifying than the playback volume data is that domestic long - form dramas are moving away from the "center of discussion". I believe many people can relate to this. In the past half - year or so, in my memory, only one drama, "The Peaceful Year", and half of "The Legend of Hidden Sea" have truly become spontaneous hot topics on social media (not just relying on bought hot searches to maintain popularity). Please note that many dramas still form local hot topics and have some die - hard fans. What I mean is that they no longer become "hot topics among the general public". As a whole, the position of the long - form drama industry in the social public opinion field has declined significantly compared to five or even ten years ago.
In short, the decline of long - form video platforms and the decline of the long - form drama industry are two sides of the same coin. Since audiences have little interest in long - form dramas, they naturally have no reason to pay for long - form video platforms. When they gradually stop opening long - form video platforms, they have even less exposure to long - form dramas... Thus, a vicious circle is formed.
The continuous decline of long - form dramas corresponds to the rise of short - form dramas: As early as 2024, the revenue scale of the domestic short - form drama industry exceeded that of the long - form drama industry. You know, at that time, AI - generated short - form dramas/AI - generated comic dramas were not even on the horizon! By the second half of 2025, the domestic short - form drama industry was all about AI, with everyone using KeLing and JiMeng. While AI production was rising strongly, long - form video platforms were in a more embarrassing position. Meanwhile, short - form drama and short - video platforms just watched silently as the entire pan - entertainment market was gradually brought under their control.
Some people say, "The rise of short - form dramas has taken away half of the user watch time of the three major long - form video platforms!" This statement may not be entirely correct, but the general direction is right. So, the question is, is this an inevitable phenomenon in the era of mobile Internet? At least, when we look at the world, we will find that this "inevitability" is not so obvious:
Netflix's performance is very good. As for how good it is specifically, readers may directly check its financial reports. Not long ago, Netflix failed to acquire Warner Bros. with a bid of $110 billion. Although it failed, it means that it has abundant cash flow and great ambition in content production.
YouTube is also developing well. Its advertising and user - paid revenues have hit new highs almost every quarter. If someone still tells you nonsense like "YouTube is being severely impacted by short - videos", you can give him a slap to wake him up from his dream and tell him, "It's 2026 now, not 2022!"
Other long - form video platforms in North America, such as Hulu, Amazon Prime Video, HBO Max, etc., are generally developing quite well. Although they are not generally independently listed, we can still roughly see their growth momentum from the financial reports of their parent companies.
Of course, short - form dramas are also developing well in North America, which has become another emerging track for domestic Internet companies going global. However, the development of short - form dramas has not brought down the North American long - form drama industry, let alone made North American long - form video platforms starve. After all, short - form/vertical - screen content and long - form/horizontal - screen content correspond to different consumption scenarios, different user mindsets, and different content types. They may rise and fall in turn, but it is difficult for one to completely replace the other. In the dimension of "IP economy" such as derivatives and secondary development, long - form dramas have irreplaceable uniqueness. After all, it is quite difficult for any short - form content to form an IP (note: it's not impossible, don't argue).
Therefore, my consistent view is that the problems faced by the domestic long - form drama industry cannot be entirely attributed to the prosperity of the short - form drama industry. As the old saying goes, 'A country must first destroy itself before others can destroy it.' For a large and mature industry, if there are no internal problems and it fails to keep up with the progress of the times, it is difficult for later - comers to overthrow it. People familiar with history should admit that even before the short - form drama industry entered the high - speed development track around 2020, the domestic long - form drama industry had already failed to keep up with user habits and mindsets, and long - form video platforms did not help them solve the problems through the "Internet thinking". I think it's not so much that the short - form drama industry "took away" the user watch time of long - form video platforms, but rather that users had long been dissatisfied, and short - form dramas took advantage of the situation to attract them away.
So, will the rising AI - generated short - form dramas/AI - generated comic dramas further ring the death knell for long - form video platforms? This time, I'm not that pessimistic. We should recognize the following points:
AI technology is neutral. It can be used to create short - form content as well as slightly longer content. The current mainstream large - scale video models already have the ability to create industrial - level long - form content. The main difficulties lie in the engineering and coordination aspects, and there are no real technical bottlenecks.
As the saying goes, "When in a desperate situation, one will seek change." In recent years, long - form video platforms have realized the importance of change. To some extent, they have shed the "old - fashioned habits" that were previously criticized (although not completely). All kinds of content platforms, including long - form video platforms, are actively exploring and trying to embrace the potential of AI - generated content.
If AI becomes the mainstream in film and television production, it may completely change the traditional drama investment and procurement models, and make the long - form drama industry shift to the "PUGC model" - which will be more beneficial to the platforms. Although this process will not be achieved overnight, and I don't think PUGC will be the only model.
There is also a very important point: Users still have a natural demand for content with high industrial standards, good completion, and long duration. In the minds of many people, the sign of a perfect weekend is still "taking a hot bath, having a cup of milk tea at hand, and watching a drama they are following on a tablet." Even the film industry, an industry that seems to be completely out of the mobile Internet era and destined to be subverted by online - born content, has not died yet (a fun fact: although the peak of domestic film box - office occurred in 2019, the box - office in 2025 was only about 20% lower in comparison).
From the so - called macro - trends, we can only conclude that "the development speed of the long - form drama industry will be slower than that of the short - form drama industry", but we cannot conclude that "the long - form drama industry will be destroyed by short - form dramas". In reality, judging from the situation from 2024 - 2025, long - form dramas have been losing ground to short - form dramas and can only defend. If this trend continues for a while, perhaps the words "watching dramas" and "following dramas" in the mouths of ordinary users will become abbreviations for "watching short - form dramas/following short - form dramas". In fact, this process is already underway...
Rather than saying that the popularization of AI production will ring the final death knell for long - form video platforms, it provides an opportunity for all kinds of content platforms to compete on a level playing field. It must be pointed out that industrial - level AI video production did not really mature until Sora 2 in October 2025 and Seedance 2 in January 2026. My estimate is that in the next 3 - 5 years, or even longer, it will still be a period for various content providers and platforms to fully understand AI production and explore the potential of AI - generated content. It is completely understandable that short - form dramas/comic dramas are in the lead because they are short, concise, and flexible in transformation, and many of their production companies come from the Internet industry rather than the traditional film and television industry.
But this by no means means that "long - form content" cannot benefit from AI production. On the contrary, from the cost perspective, long - form content will benefit even more. Since short - form dramas are originally a low - cost content form, the improvement from "low cost" to "even lower cost" is limited, although there is still progress (note: I'm not saying there is no progress, don't argue). Traditional long - form content, which is regarded as resource - intensive and requiring high industrial standards, may, for the first time, become a content form that "ordinary creators can afford" with the rise of AI. In half a year, when the next - generation large - scale video model matures, this trend will be even more obvious.
I can't wait to see this day come. The biggest beneficiaries of this scenario will be the audience. One day, they will be able to see unprecedented and dazzling content.
This article is from the WeChat public account "Internet Thief Group". Author: Pei Pei, the leader of Thief Group. It is published by 36Kr with authorization.