Hongguo’s 168M DAU and the AI Micro-Drama Shift: What It Means for Teams Producing Media at Scale in 2026
The economics of scaled media output shifted in 2026, and the platform that moved them is one most Western production teams have never opened. Hongguo Short Drama, ByteDance’s short-drama app, was estimated at 168 million daily active users in July 2026 by the data firm QuestMobile, a figure widely reported but not confirmed by the platform itself. Even read as an estimate, it points to more daily opens than any single Chinese long-video service. Daily active users, or DAU, counts the people who open an app on a given day, and 168 million of them is a supply signal, not just an audience one.
Micro-drama is the format behind that number: serialized vertical video, episodes of one to three minutes, built for phone viewing and released in batches. It is also the format AI micro-drama pipelines now fill at volume, which is why the same year brought Netflix’s own quarterly filing put the price at about $587 million in cash for an AI filmmaking studio.
Two halves of one story: a platform proving demand at scale, and a major buyer converting that demand into capital. What follows is a dated, source-attributed read of what changes for teams shipping media at volume.
What the Data Shows: Hongguo Short Drama’s Estimated DAU Topped China’s Four Biggest Long-Video Platforms
In July 2026, Hongguo reached 168 million daily active users, up 107 percent year over year from roughly 100 million in January, according to QuestMobile monitoring data relayed in a Nomura research note. That figure is not a single outlet’s claim: 36Kr’s English edition and the South China Morning Post both report the same QuestMobile-for-July numbers, while Caixin documents Hongguo’s operator, Beijing Bimo Liuxiang Technology, as a ByteDance (Douyin) subsidiary. DAU here means users who opened the app that day, QuestMobile’s caliber, not verified viewing sessions.
The same research summary put the combined DAU of iQIYI, Tencent Video, Youku and Mango TV at about 155 million, with Tencent Video near 50 million and iQIYI near 45 million (Sohu, citing QuestMobile; 36Kr, in English). Both sides of the comparison trace to one measurement house, QuestMobile, so the ratio is internally consistent even though it is not a third-party audited number. Hongguo also logged about 125 minutes a day per user in February 2026, and total monthly usage duration rose 157 percent year over year in July (36Kr, citing Nomura and QuestMobile).
⚠️ Warning: A substantial share of those opens are check-in tasks and watching ads for coins, so incentivized DAU is not equivalent to self-directed viewing (Sina Finance). The 168 million figure traces back to one measurement house, QuestMobile, and is a point-in-time snapshot: a later monthly report put Hongguo at roughly 186 million DAU in August. Treat both as research estimates rather than audited platform disclosures.
Why It’s Happening: AI Micro-Drama Supply Became the Volume Driver
An AI comic-drama is a short-form series whose visuals are generated rather than filmed, and on Hongguo it is now the volume engine rather than a novelty category. AI comic dramas took seven of the top ten slots on Hongguo’s chart in the second week of July 2026, led by the top title at 89.19M heat against 67.11M for second place.
Live-action titles fell from all ten slots in March to four by July, as of July 20, 2026. DataEye’s July ranking, which orders titles by peak heat, totalled 5.911B across listed titles, with all of its top 13 slots held by 3D comic dramas.
Read the share carefully, because it depends on the chart. The weekly reading is seven of ten; a September retrospective repeats that same seven-of-ten figure, while broader top-ten claims put AI-generated micro-dramas at eight of ten. Classification also varies by source, since some charts count AI-assisted live-action alongside fully generated titles.
The Free-to-Watch Plus Ad-Revenue-Share Model That Makes High-Volume Iteration Rational
Hongguo’s model is free to watch plus ad revenue share: users pay nothing and the platform splits ad revenue with producers, and payouts are tied to play volume and effective viewing time (Sina Finance, 2026-09-09). That structure rewards volume over polish, and the platform reinforced it with an RMB 1.5 billion support budget for live-action short dramas (36Kr, 2026-09-09).
The economics are harsher than the volume suggests. Even as about 90 percent of short-drama projects lost money across more than 360,000 titles released or renamed in H1 2026, one operator put whole-business net margin below 1 percent (Sina Finance, 2026-09-09). More than 95 percent of the 128,000 mini-dramas released in Q1 2026 were AI-generated, yet only 1,055 of more than 221,900 new AI short dramas passed 100 million views, a 0.47 percent rate (Yicai Global, 2026-08-03). Per-10,000-play revenue fell from 30 to 100 yuan down to 5 to 10 yuan (Sina Finance, 2026-09-09), while live-action releases fell to 1,629 in June from 1,934 at the start of the year (Yicai Global, 2026-08-03).
⚠️ Warning: A 0.47 percent hit rate means roughly 1 in 200 AI titles clears 100 million views. Volume only pays if your per-episode cost and testing cadence are built for that failure rate.
The Attention Shift: ByteDance Now Takes 40.9% of Chinese User Time
The DAU milestone is a symptom, not the story. The story is where Chinese attention actually sits, and how quickly it moved.
In July 2026, ByteDance’s five apps accounted for 40.9% of total usage duration across China’s top 50 apps, up 7.3 percentage points from 33.6% a year earlier, while Tencent’s ecosystem held 29.1% (36Kr, citing Nomura’s QuestMobile analysis, September 2026). At the single-app level, Douyin’s usage-duration share reached 19.4%, ahead of WeChat at 18.8%, with Douyin up 29% year over year against WeChat’s 3% DAU growth.
The intensity gap matters more than the ranking. Douyin users averaged roughly two hours a day, 112 to 120 minutes against WeChat’s 82 to 85, on 714 million versus 931 million DAU (TechNode, citing Nomura, September 2026).
One caveat on the numbers: June 2026 data put ByteDance at 40.1% and Tencent at 29.7%, while the July figures above read 40.9% and 29.1%. Both sets trace to the same upstream Nomura analysis of QuestMobile data, so treat the direction as reliable and the exact decimals as provisional. Independent English-language coverage from TechNode and 36Kr reports the same trajectory.
The Global Picture: Hollywood Moved From Statements to Capital in 2026
Western studios stopped announcing AI ambitions and started paying for them. Netflix’s own quarterly filing put the price at about $587 million in cash for Ben Affleck’s InterPositive, an acquisition completed in March 2026. The disclosure came in the company’s Q2 2026 Form 10-Q filed with the SEC, a primary-source figure corroborated by Variety, TechCrunch and Deadline. That is AI film and TV investment at studio scale, not a pilot program.
The pattern repeats across buyers. Netflix says more than 300 of its programs use AI in production, Lionsgate took an equity interest in Runway as part of a June 2026 expansion, and Google DeepMind put $75 million into A24 alongside research partnerships with filmmakers Doug Liman and Darren Aronofsky.
Key Takeaway: The Western shift is now balance-sheet money, not press releases. Netflix, Lionsgate, and Google DeepMind are funding AI production capability directly, which moves the question from “will studios adopt this” to “how fast can teams staff and ship against it.”
The supply side is formalizing too. Variety’s September 2026 survey of ten companies embedding AI into production includes Amazon MGM Studios, Black Forest Labs, and Particle 6, whose Tilly Norwood-led feature “Misaligned” began a 12-month production cycle in September 2026. Lionsgate named its first chief AI officer in February 2026, and Black Forest Labs was cited by Martin Scorsese for storyboarding. Pre-visualization, the planning stage where scenes are sketched before shooting, and post-production, the editing and effects work after filming, are now funded line items rather than experiments.
What This Means for Teams Producing Media at Scale
The operator takeaway is blunt: throughput, testing cadence and cost per episode now decide outcomes more than per-title polish. Four shifts follow from that.
Pipeline throughput. When a studio is shipping 30 or more episodes a month, the constraint stops being production capacity and becomes review and release cadence. Approval queues, not cameras, set the schedule.
Testing volume. A 0.47 percent hit rate means the portfolio, not the title, is the unit of planning (KR-Asia, 2026). You budget for a slate and expect most of it to miss.
Cost structure. Per-10,000-play revenue falling from RMB 30 to 100 down to RMB 5 to 10 changes what a viable episode budget looks like. Episodes that made sense at the old rate do not survive the new one.
QA and provenance. With AI-generated supply at 95 percent of Q1 2026 releases, classification and quality gates become a distribution problem, not only a creative one.
One honest limitation: the 168 million figure is a QuestMobile estimate relayed through Nomura, not a platform disclosure, and the long-video comparison comes from the same upstream house.
Pro Tip: Set your kill threshold before the slate ships. Decide in advance how many episodes a series gets to find an audience, and hold to it when the data comes back flat.
What to Do Now
Start with instrumentation, not reaction. Measure cost-per-finished-episode and per-title hit rate on your own pipeline before you change anything in response to a market number.
This month, run a small AI-assisted slate as a portfolio test rather than a flagship bet, and write the kill threshold down before the first episode ships. This quarter, build provenance and classification metadata into your release process so AI-assisted titles can be labeled, tracked and compared. Keep tracking attention metrics on the platform where your audience actually is.
Key Takeaway: Instrument first, test small with a pre-agreed kill threshold, label AI-assisted titles, and follow your own audience’s platform. Do not rebuild a production strategy around a single-source DAU figure, and do not assume a Chinese-market supply shift transfers to your market on the same timeline.
The Bigger Picture
Both stories are one story: attention and production capacity are decoupling from legacy distribution. Viewing time keeps migrating to short-form vertical, yet long video is not dying so much as moving screens. iQIYI’s TV app drew more daily active devices than its mobile app, and its smart-TV audience averaged 256 minutes a day in August 2026 (Korea Herald, retrieved 2026-09-18). Meanwhile, AI film and TV investment moved from demonstration to slate, with Particle 6’s Tilly Norwood-led feature entering a 12-month production cycle. The open question for 2026: does that capital produce a durable hit rate, or a second wave of write-downs?
Key Takeaway: The Hongguo milestone and Hollywood’s AI spending are the same event. Attention and production capacity are both leaving legacy distribution behind, and the winners will be the teams that can supply volume without losing quality control.
Frequently Asked Questions
Is Hongguo Short Drama really bigger than Tencent Video, iQIYI, Youku and Mango TV combined?
Yes, on the QuestMobile estimate. Hongguo was put at 168 million DAU in July 2026, while China’s four biggest long-video platforms together sat at roughly 155 million in the same data set. Three caveats matter. The comparison pits a single short-video app against four long-video apps, not a like-for-like category. Both figures trace to the same measurement house. And neither is a platform disclosure, so treat the gap as a strong directional signal rather than an audited number.
How much of Hongguo’s content is AI-generated?
AI micro-drama dominates its charts. Seven of the top ten titles in a recent weekly ranking were AI-generated, and broader tallies put the share near eight in ten. Classification varies by source, because “AI-generated” can mean a fully synthetic pipeline or a human script with AI-assisted rendering. The direction is consistent even where the exact count is not.
Does the free-to-watch ad revenue share model actually make money for producers?
For most, no. Roughly 90% of AI micro-drama producers lose money, and operators report net margins below 1%. The hit rate is about 0.47%, so the model works as a portfolio bet: a small number of titles carry the returns for everyone else.
What does this mean for non-Chinese media teams?
The transferable lesson is throughput and portfolio planning, not the exact format. Teams that can ship many low-cost episodes and kill weak ones fast are better positioned than teams optimizing a single title. Market timelines differ, so treat the reported Chinese numbers, which come from a single measurement house, as a leading indicator rather than a schedule.
Conclusion
- Hongguo Short Drama was estimated at 168 million daily active users in July 2026, a QuestMobile figure relayed via Nomura and reported by 36Kr and SCMP, but not a platform disclosure, so treat the exact number as directional.
- AI micro-drama is now a supply engine, not a novelty: AI comic dramas took seven of the top ten slots on the platform’s charts.
- The economics reward volume. Episodes are free to watch, paid by ad revenue share, which makes a low hit rate survivable as long as iteration stays cheap.
- Attention is consolidating fast: ByteDance’s five apps now account for 40.9% of time spent, so distribution leverage matters as much as production quality.
- Hollywood has moved from statements to capital, and Netflix’s own SEC Form 10-Q put the price at about $587 million in cash for Ben Affleck’s InterPositive.
If you want to follow this shift as it develops, the useful habit is to track supply-side signals, meaning chart composition, release cadence and platform payout terms, rather than the headline user numbers alone.
All figures in this article trace to publicly available sources — QuestMobile via Nomura, 36Kr, SCMP, Sina Finance, Yicai Global, TechNode, and Netflix’s SEC Form 10-Q — all retrieved September 18, 2026. Hongguo/Hongguo Short Drama DAU figures are QuestMobile estimates relayed via Nomura, not platform disclosures; treat exact numbers as directional.
