AI News

Tencent's Fourth Foray into Short Dramas: How Can WeChat Entry "Green Bubble" Challenge Hongguo

HuxiuUpdated Nov 16 18:51

AI summary

In October 2026, Tencent launched its fourth foray into short dramas through the official WeChat entry "Green Bubble," adopting a free-to-watch, ad-supported IAA model without producing its own content, directly challenging Hongguo Short Drama, which has 168 million daily active users. On October 14, Green Bubble released about 30 new titles in a single day, while Hongguo released more than 230, with a stark gap in likes for top titles. The article also notes that the full-scope short-drama market is projected to exceed 121 billion yuan in 2026, yet only 1.3% of AI short dramas launched in the first half of the year recouped their costs, and more than 40% of production companies disappeared within six months.

Why read this

AI score 57

The article places Tencent's fourth short-drama attempt against Hongguo's business chain and the cost curve of AI video models, making clear the misalignment among entry points, content, and the model ecosystem.

Topics

Tags

Body · AI translation

[[**In October 2026, Tencent made its fourth push into short dramas, launching the WeChat entry point "Green Bubble," adopting a free-plus-ads (IAA) model, producing no content itself and only providing an entry point and traffic, directly challenging Hongguo Short Drama with its 168 million daily active users, but facing a content generation gap and the awkwardness of social relationship chains.**

**Key Points**

**1. After four pushes, Tencent shifts from making "premium" content to "free-plus-ads"**

Tencent's first three short drama attempts (Shifen Theater, Mars Short Drama Club, WeChat short drama mini program) all failed. In October 2026, "Green Bubble" launched as an official WeChat entry point, producing no content itself and relying entirely on third-party short dramas, fully replicating its rival's IAA model.

**2. Green Bubble's biggest advantage is the WeChat social chain, but its content gap with Hongguo is obvious**

Green Bubble requires no download and directly shares relationship chains and regional trends. But on October 14, it released only about 30 new titles in a single day, far below Hongguo's 230; top series received single-digit likes, while Hongguo's reach the millions. Users' "short drama shame" leaves little motivation for social recommendations.

**3. The short drama market is booming, but 98.7% of projects are unprofitable, and producers are being cleared out en masse**

In 2026, the full-scope short drama market is expected to exceed 121 billion, but the industry's core profitability metric, "revenue per ten thousand views," has plunged nearly 90%, and ROI for small and medium teams is generally below 0.8. Of the AI short dramas launched in the first half of the year, only 1.3% recouped their costs, and more than 40% of production companies disappeared within six months.

**4. AI video models find their first large-scale profitable scenario through short dramas**

This article is from the WeChat public account: Xie Pu Notes, author: Xie Pu

In July 2026, Hongguo Short Drama had 168 million daily active users, up 107% year on year. Over the same period, Tencent Video had about 50 million, iQiyi about 45 million, and Youku and Mango about 30 million each—together, the four total 155 million.

A single short drama app beat the entire long video industry.

Note that Hongguo Short Drama only launched independently in August 2023, and it took just two and a half years to surpass 100 million daily active users. Back then, Tencent Video took more than six years to climb to that level, and iQiyi took four and a half years.

Four months later, something called "Green Bubble" quietly grew inside WeChat. In mid-October, the old mini program entry point called "Short Drama" went offline, and it took over completely.

This is Tencent's fourth push into short dramas. This time, the playbook is different from the previous three.

English text is being prepared.

Strictly speaking, it's not that Tencent never fought in short dramas—it has fought for four years. It changed forms three times.

English text is being prepared.

In January 2025, "Mars Short Drama Club" launched on WeChat, later renamed "Mars Drama Watching." An independent mini program that never obtained WeChat's core traffic and remained stuck at the edge as a trial.

In June 2025, WeChat's official "Short Drama" mini program launched. In early 2026, WeChat revamped and raised the entry point level for "Short Drama"—in WeChat's world, this counts as taking it seriously.

In July 2026, "Short Drama" was renamed "Green Bubble Short Drama," entered gray testing in August, debuted at the end of September, and replaced the old entry point in October. During the gray testing period, one sample emerged: the AI simulated-human short drama "Best Actress Yaya," first released on August 21, surpassed hundreds of millions of views in three days, with peak concurrent online users exceeding 2 million.

Channel page, independent mini program, official entry point embedded in the relationship chain. Each form lighter than the last, each posture lower than the last.

What also changed was Pony Ma's tone. At the annual meeting in early 2025, he said Tencent would not make low-quality short dramas and would focus on investing in premium short dramas. At the annual meeting on January 26, 2026, his attitude toward short dramas was: Tencent Video was hit by short dramas last year, and in addition to persisting in high-quality long video investment, next it "also needs to make some ad-driven products."

Translation: the premium battle was fought, and not won. Now, fight again using the opponent's methods.

English text is being prepared.

What is Green Bubble? Broken down, it has almost nothing.

This is the most counterintuitive part of this war: a product directly challenging Hongguo does not produce a single drama itself.

For more than a decade, Tencent has never fought a content war like this. In the long video era, it competed with iQiyi on original productions and copyrights; in the online literature era, it held China Literature. This time, it removed "content" from the bet entirely.

In fact, short drama users inside WeChat have long existed. As of February 2026, short drama mini programs within the WeChat ecosystem had reached 427 million monthly active users, more than the total of all independent short drama apps—only this traffic was scattered across dozens of mini programs each fighting for itself.

What Green Bubble does is gather them into one official entry point.

So for the first time, Tencent packaged its entire social relationship chain and threw it into the short drama battlefield. The homepage features a "Friends Are Watching" recommendation slot; what friends have watched helps shape your content feed; "Regional Trends" directly presents "TOP20 that Beijing friends are all watching"; floating playback in a small window means when a WeChat message arrives, the drama shrinks into a small window so you can reply while watching.

The supporting money is also in place: in Q4 2026, premium new dramas participating in co-distribution can receive up to 100% revenue incentives, including 90% cash sharing and 10% advertising credit matching. On the producer side, stocks have already risen in anticipation—since October, Zhangyue Technology hit the daily limit, Guomai Culture touched a 20% daily limit intraday, and Chinese Online rose in sympathy. The logic is written in the candlestick chart: one more buyer means one less chance of being squeezed on price.

In 2025, the AI animated micro-drama market reached 18.98 billion yuan, up 276.3% year on year; in the first five months of 2026, AI dramas and comic dramas had already hit 22 billion yuan, with a full-year forecast of 40 billion, a further rise of 138%. And that is only the view through the narrow gate—counting live-action micro-dramas as well, the full-scope market is expected to exceed 121 billion yuan in 2026. In 2024, the micro-drama market overall stood at 50.44 billion yuan, surpassing annual box office for the first time; in 2025 it reached 67.79 billion, against 51.8 billion in box office, widening the gap to 16 billion. If one takes only the narrow category of "AI comic dramas," the full-year figure is about 24 billion—different statistical scopes produce conflicting numbers, but there is no disagreement on the trend: a curve that covers in one year the distance others took five.

Beyond the consumption side, the supply-side numbers are even crazier. In Q1 2026, the industry released about 128,000 micro-dramas, of which 122,000 were AI micro-dramas, accounting for more than 95%—a single quarter's releases were close to four times the whole of 2025. At the peak, a new AI short drama went live almost every 20 seconds. As of April 2026, there were 22,000 AI short-drama-related companies, with 5,465 newly registered in 2025 alone.

Behind the frenzy, boom and bust coexist. The turning point is the collapse of the cost curve.

In March 2026, ByteDance released the video model Seedance 2.0, with commercial pricing equivalent to about 1 yuan per second—14 yuan to generate a 15-second 720p video. At the end of July, Seedance 2.5, Kuaishou Kling 3.0, MiniMax H3, and Alibaba Wan 3.0 launched in quick succession, with multi-shot consistency, audio-visual sync, and character stability crossing the "mass-producible" threshold, and MiniMax H3's 2K generation directly hitting 0.8 yuan per second.

The cheapness of creation costs was quickly converted into collapsing market quotes: a one-minute AI short drama cost about 5,000 yuan in the second half of 2025, and now costs a few hundred yuan; the industry has even formed an unspoken "four-tier pricing," with the lowest tier under 50 yuan per minute. By comparison, a traditional live-action short drama costs 500,000 to 800,000 yuan per title to produce, versus 80,000 to 150,000 for an AI short drama—work that once took a 20-person team three months is now finished by three to five people in a week.

The barrier to making content has collapsed, and the number of people rushing in has doubled. The feast was laid on the table, and soon some discovered that they were not the guests but the dish.

English text is being prepared.

Whether short dramas can keep thriving like short videos is full of uncertainty, but the problem seems to be the old one—content has gotten cheap, yet the money has not stayed in the hands of the people making it.

The value of the short-drama business is being siphoned off at both ends and thinned out in the middle.

Upstream is IP. In March 2026, Douyin's short-drama copyright center launched a comic-drama IP library, opening adaptation rights to more than 500,000 original Fanqie novels in one go—all of these novels have been validated in the free-reading market, making the cost of trial and error in topic selection almost zero. The effect was immediate: of the 664 annual hit short dramas across the web in 2025, 413 were adapted from Fanqie IPs, accounting for 62%.

Downstream is the gateway. Hongguo holds the revenue-sharing coefficients and the power to allocate traffic; according to multiple suppliers who spoke to the media, its related revenue in 2025 already exceeded 15 billion yuan, though the company has never disclosed it officially. Producers have no bargaining power here—in the industry's words, they can only passively accept the revenue-sharing coefficient.

In the middle are the producers, and their books are the coldest. The industry's core profitability metric, "revenue per 10,000 views," fell from 30 to 100 yuan in the second half of 2025 to 5 to 10 yuan in 2026, a drop of nearly 90%. Mature traffic-buying teams hold ROI steady at 1.03 to 1.07—for every 100 yuan spent on user acquisition, 103 to 107 yuan comes back, a book profit of three to five points, and one slip means a loss; small and mid-sized teams are generally below 0.8, losing more the more they spend. Taking the industry-recognized break-even line of 50 million views, only 1.3% of the AI short dramas launched in the first half of the year recouped their costs.

This is the most piercing number in this article: 98.7% of projects fail to reach break-even within six months of launch; on average, only 1 in every 77 new dramas recoups its cost. Qichacha's data is even more blunt: in Q1 2026, active AI comic-drama producers fell from 1,216 to 698, with 42% of companies disappearing within half a year. The industry's self-mockery says it all: the people making content lose money, while the people selling tools and traffic make money.

And the biggest seller of traffic at that table is Hongguo.

In 1992, Acer founder Stan Shih proposed the famous "smile curve" theory; 34 years later, it still applies today. It truly does not deceive me.

English text is being prepared.

The mountain opposite Green Bubble is higher than most people think.

In January 2026, Hongguo's daily active users broke 100 million; in May, monthly active users reached 356 million, up 79.1% year on year; average daily time spent per user was 125 minutes, surpassing Douyin and WeChat, and in August it climbed to 156.9 minutes.

A summary of exchanges between Nomura Securities and industry experts judged that Hongguo is expected to meet its full-year core revenue target of 42 billion to 45 billion yuan.

More noteworthy is the user structure. Users aged 50 and above number 40.745 million, accounting for 13.4% of total users; while the entire short-drama app sector has a total of 46.53 million silver-haired monthly active users, Hongguo alone accounts for nearly 90%. Grandpa's and grandma's phone time is almost entirely wrapped up by it.

In fact, Hongguo is not just a product—it is not fighting alone, but is a self-turning business chain: Fanqie novels supply IP, low-cost adaptation and production follow, Douyin clips drive traffic, Hongguo hosts the full episodes and monetizes through ads, then shares revenue with producers.

IP, traffic, and money circulate within one system. In 2026, its guaranteed support budget for live-action short dramas exceeds 1.5 billion yuan, with the per-title guarantee up 60% year over year. For small and medium-sized producers nationwide, Hongguo's revenue sharing and payments are their cash flow. China Aviation Securities sums it up in eight words: one dominant player, hard to shake in the short term.

In other words, what Lüpaopao really needs to win over is not so much users as the hearts of producers: to tell them that besides Hongguo, there is a second buyer in this world.

English text is being prepared.

On Tencent's books, this battle looks like a sure win. This battle is not Weishi; it is the glory of Channels reborn.

Every single daily active user at Hongguo has to be bought with ad spend; Lüpaopao's daily active users grow inside a chat app with roughly 1.4 billion monthly active users—no download, no registration, just tap and watch.

But short dramas are not Channels.

Channels could break out through the social chain because forwarding a video is self-expression; watching "My Husband Went Bankrupt, the Vase Wife's Identity Can No Longer Be Hidden" is not.

In mid-October, this drama was simultaneously ranked sixth on Hongguo's hot search list and featured on Lüpaopao's homepage recommendation slot, while the buzz on social platforms was uniformly "I don't dare log in, too embarrassed to like." The new term "short-drama shame" was coined in these very days. Lüpaopao's "Friends Are Watching" only shows how many friends have watched, not who; but once you tap the heart, it will—like Channels—show exactly which friend recommended it.

Zhang Yi, CEO of iiMedia Research, pierced through this thin veil: short-drama viewing preferences are not necessarily content users are willing to show to acquaintances; mini-programs are accessible, but "activation is relatively weak," and they cannot cultivate high-frequency habits. Hongguo users can actively reject a drama, but WeChat is a giant recommendation arena—when homogenized domineering-CEO thrillers are pushed to circles of acquaintances week after week, what is consumed is the restraint WeChat has accumulated over more than a decade.

The content ledger is even colder. On October 14, Hongguo released more than 230 new titles in a single day, while Lüpaopao released about 30; the top drama on Lüpaopao's overall chart had 10 likes and 3 shares on its first episode, while likes on Hongguo's hit chart are in the millions.

One analyst's judgment is blunt: Lüpaopao has a generational gap in content compared with rivals, and relying solely on WeChat to pump blood will not let it break out of the red ocean.

Nomura provides a bigger backdrop: in August 2026, ByteDance's apps took 41.8% of Chinese netizens' usage time, with Hongguo accounting for 6%, a year-over-year surge of 182%; Tencent's apps fell back to 28.6%.

Tencent cannot afford to lose this battle.

So Lüpaopao is less an attack by Tencent than an admission by Tencent: an admission that the battlefield ByteDance chose is the right one, an admission that content is no longer scarce, an admission that free plus advertising is the main artery of this industry.

But the download cost saved by mini-programs cannot buy a generational gap in content; pouring homogenized thrillers into the chain of acquaintance relationships is precisely what WeChat has most carefully avoided for over a decade. What this battle ultimately comes down to is not who has more traffic, but whether Tencent dares to make that difficult decision: not pushing "domineering CEOs" to friends.

The death of Sora is the most easily overlooked background note in this short-drama war.

On March 24, 2026, OpenAI announced the shutdown of Sora—the AI video model that two years earlier had stunned the world with a "stroll through the streets of Tokyo" and surpassed a million downloads in its first few days online. The exit timetable was clean and crisp: web and app services shut down on April 26, and the API closed on September 24.

The reason is not dignified, but it is honest: video generation is the most compute-hungry task in AI, Sora's daily compute cost was in the millions of dollars, while its cumulative consumer revenue since launch was only about $1.4 million. Users' novelty also lasted only two weeks—

In the first two weeks after launch, we made over a hundred videos; two weeks later, we never opened it again.

In the same year, on the other side of the planet, video models found their first soil capable of monetizing at scale—Chinese short dramas. The logic is simple and crude: models bill by the second, while the entire industry releases 120,000 short dramas a quarter, with each title drawing dozens of assets over and over. The bottomless pit of compute in American labs became, here, a means of production with a clearly marked price. Production quotes fell from 4,000 to 5,000 yuan per minute in 2025 to 300 to 500 yuan in 2026—the 90% price gap that was cut away is the market the models ate.

There are three players at the table, conveniently representing three ecosystems.

ByteDance is the most complete: Seedance is embedded in Jimeng, Jimeng connects to Jianying, and Jianying connects to Douyin and Hongguo—generation, editing, distribution, and monetization all turn in one chain of its own. In the industry, for those making AI short dramas, "Jimeng runs the whole film, Jianying does the voiceover" has already become the default workflow—this playbook can be summed up in five words: swap gun-making for target practice.

Alibaba takes another path: in August 2026, Wan 3.0 launched, able to directly read PPTs to generate product videos, turn white-background images into dynamic displays, turn portrait images into virtual models, with 480P as low as 0.3 yuan per second—it targets the asset pipelines of millions of e-commerce merchants, stitching the model into merchants' workbenches before the big promotion season.

Tencent is the most awkward: the Hunyuan video model is not bad in Chinese context, but its buzz and user scale are not in the same league as the other two, and industry reviews describe it as "conservative in iteration" and "unable to feel the ecosystem advantage." AI models are not like content—content can be bought, and all of Lüpaopao's dramas come from third parties; models, however, can only be run out in one's own scenarios. Every time a user draws a card, data and revenue come in at the same time.

Lüpaopao therefore carries an additional meaning that few mention: it is not just a short-drama entrance, but also the soil Tencent has found for its own AI. Hongguo fattened Seedance; WeChat may well be able to keep Hunyuan alive.

Source

Huxiu