AI short dramas target 40 billion yuan: Don't rush to calculate hit rates

Last week, I helped a friend review a business plan for an AI short drama company. The cover page claimed annual revenue could hit 40 billion yuan and production costs had dropped by about 90%. I only asked who ultimately makes money from this, and the other party froze. 40 billion looks more like a revenue caliber frenzy, still far from profit caliber. Cost reduction has indeed happened, but the explosion in capacity dilutes traffic per work to very thin levels. The scariest thing in valuation is mistaking "can produce" for "can make money." To see where the market ceiling is and what the competitive landscape looks like, we need to focus on who holds the gates for distribution, ad spending, IP, and overseas expansion. Whether models can generate content is just the baseline.

According to DataEye metrics, the domestic AI short drama market exceeded 22 billion yuan in Jan-May 2026, with annual revenue expected to hit 40 billion yuan, a growth of about 138% compared to 16.8 billion yuan in 2025. In the first half of the year, 367,000 micro-dramas were launched online, with AI short dramas accounting for over 74%. Traditional short dramas require about 15 people and 4+ weeks; AI short dramas average 3-5 people and about 10 days. Standardized production quotes have dropped from 5,000 yuan per minute to a few hundred yuan, approaching the cost line; custom short dramas can still maintain 10,000-20,000 yuan per minute.

This change goes beyond simple cost reduction. Short dramas are shifting from small workshop films to fast-moving consumer content goods. Once the capacity threshold disappears, the dimensions of competition change. Previously, it was about who could shoot it; now it's about who can be seen and who can recoup ad acquisition costs. Public reports say shooting short dramas is as easy as posting on WeChat Moments, but only 1 out of every 77 new dramas breaks even, and 98.7% haven't touched the break-even line within six months. If revenue growth accompanies extremely low ROI, the market values this more like a traffic casino than a content company.

Cost reduction doesn't automatically translate to profit uplift

I wrote an article earlier reflecting on AI costs, emphasizing that productivity markets calculate accounts first. Looking at AI short dramas now, it's the same logic: costs drop 90%, but profits won't automatically rise 90%. Average production cost for premium live-action paid short dramas is about 1.5 million yuan, while vertical screen feed AI pseudo-live-action short dramas can be controlled within 200,000 yuan. But production cost is only part of total investment; ad spending, platform commissions, channel distribution, and copyright procurement don't decrease, so profits are still sliced away. User attention has become the most expensive raw material.

From the industry chain perspective, upstream models and tools will rapidly commoditize. Generation, continuation writing, and character consistency are selling points today but will likely become infrastructure in six months. Midstream production companies that only know how to use AI will resemble low-margin outsourcing. Premiums lie in customization, high-quality narrative, ad creative libraries, and overseas localization. Downstream platforms are stronger because short drama consumption paths rely heavily on recommendations, search, paywall triggers, and ad feedback loops. The closer the platform is to the user, the closer it is to the money.

I'd divide the competitive landscape into three layers. Layer one is traffic entry points, controlling distribution and transaction links, most likely to have stable profit pools; layer two is IP and copyright holders, as supply explodes, users will concentrate on familiar brands and continuous worldviews; layer three is production tools and content factories, which are the most numerous and face the fastest elimination. The film and TV sector will be repriced due to AI long-form dramas getting broadcast slots, such as Journey to the West landing on Hunan TV's prime time. But stock prices react to narratives first, then must return to single-work ROI, inventory turnover, platform revenue sharing, and ad conversion.

Overseas expansion is another variable. The overseas AI short drama market size is projected to reach $650 million in 2026, growing about 6x YoY; the overseas market for AI dramas/comics is expected to exceed $4 billion, a 390% increase from 2025. Domestic capacity is too dense, leaving windows for overseas platforms, multilingual localization, and cultural adaptation. Overseas expansion involves localization, pushing competition from "who generates faster" to "who understands local payment habits, ad channels, and compliance boundaries." The Micro-Drama Development Management Measures are about to take effect, meaning the industry will eventually shift from wild launches to filing, copyright, and content quality standards. For top platforms, compliance is a barrier; for long-tail workshops, compliance is a cost.

Looking ahead 12 months, I believe 40 billion in revenue will continue to grow, but profits won't rise in sync. Standardized mass production will further intensify price wars, with some teams surviving on subsidies and platform activities; premium customization, IP adaptations, and overseas localization will retain margins; platforms mainly view AI as a cost-reduction tool, not necessarily a profit source. In terms of valuation, the market will shift from user numbers, launch volume, and play counts to single-work break-even rate, re-watch rate, LTV/CAC, and copyright asset turnover. If a company only talks about capacity and not break-even, I'd likely give it a discount. The market ceiling mainly depends on attention distribution; generative models are just the foundation. The competitive landscape mainly depends on who can turn cost reductions into others losing money while collecting rent themselves; tool capability is just the ticket to entry.