Viral AI Short ‘Huo Qubing’ Director Pushes Back: 500M Views Not Verifiable, RMB 3,000 Was Compute Only

The key angle

Director Yang Hanhan (杨涵涵), whose AI-generated short film “Huo Qubing” (霍去病) surged across Chinese social media, has clarified headline-grabbing claims about its reach and cost. It has been reported that the film amassed more than 500 million views, and that it was made for just RMB 3,000 (about $420). Not so simple, he says. According to IT Home (IT之家), Yang noted he cannot independently tally the view count and that the RMB 3,000 figure refers only to compute spend, not total production costs.

Clarifications

Yang said the “500 million views” figure originated from Taiwanese broadcaster EBC News (东森新闻) and was cited without rigorous verification. He also disputed viral assertions that the project was an “80-episode” series: the team produced just one short, cut into two versions—one running a little over four minutes and another about six minutes. Other circulating clips are not from his team, he added. The film was made by three people—a writer-director, an AI production lead, and a music/sound editor—over 48 hours, while his broader studio comprises roughly 20 specialists focused on AI film production.

Why it matters

Was this a shoestring blockbuster? Not quite. China’s short-video ecosystem—anchored by platforms like Douyin (抖音), Kuaishou (快手), and Bilibili (哔哩哔哩)—is awash in AI-assisted “short dramas,” where splashy claims about ultra-low budgets can go viral. But compute is only one line item. Labor, rights management, sound, and distribution still matter. “Huo Qubing,” which references a famed Han dynasty general, also taps into a popular historical-epic vein that often travels fast online, fueling copycats and metric inflation.

The bigger picture

The episode underscores two currents in China’s AI media boom. First, production efficiency is increasingly prized as U.S. export controls on advanced chips constrain access to top-end AI hardware, sharpening the focus on cost-effective pipelines. Second, regulators have tightened rules on “deep synthesis” content, requiring labeling and curbing misuse—pressure that raises the stakes for accuracy in promotion. Against that backdrop, Yang’s recalibration—acknowledging unverifiable numbers and narrowing the cost claim—offers a reality check amid accelerating hype.