Once-bustling short drama production bases such as Hengdian and Zhengzhou are now shrouded in an unprecedented silence. With the rapid advancement of AI technology, the short drama industry is undergoing a harsh "reorganization." According to recent industry observations, several major platforms are actively promoting a new production model featuring "live actors as leads and AI characters as supporting roles." Local actors who once relied on background roles or extra appearances to make a living have become the first "sacrifices" in the face of AI technological impact.
Currently, the once-busy notice groups at short drama production bases have completely gone silent. It's not just mid-level actors and background actors facing mass unemployment; even top-tier short drama actors who previously earned tens of thousands of yuan per day have seen their pay cuts by half. This drastic decline stems from a fundamental shift in the production logic of producers.
Data shows that in January 2026, AI-generated human-like short dramas accounted for 38% of the top 100 short drama lists, up from 7% the previous year.
Cost Killer: A traditional high-quality live-action short drama typically costs between 1.5 million and 3 million yuan, while a similarly high-quality AI short drama can be produced for less than 200,000 yuan.
Ultra Low Unit Cost: Relying on the latest commercial technologies from ByteDance, the cost per episode of AI short dramas can be reduced to below 500 yuan, just a fraction of the cost of live-action productions.
The cost of live-action short dramas is limited by locations, labor, and high salaries, leaving little room for reduction. In contrast, AI short dramas fully benefit from the declining cost of computing power.
Efficiency First: A leading short drama platform expects to produce 150 episodes per month by the end of March 2026, a growth rate that real-world filming-based live-action crews cannot imagine.
Blue Ocean Market: It is expected that the market size of AI short dramas will reach 24 billion yuan in 2026, with user scale potentially reaching 280 million.
