What happened
A production group named Can Diep, based in Xinxiang, China, has achieved significant viewership for its AI-generated short film series, 'Manier les armes à travers le temps sous la dynastie Song.' According to reporting by Vietnam.vn citing 163.com, the series, which blends time-travel narratives with modern military elements, reached 6 billion views by November 3, 2026. The production team, consisting of approximately 30 members from diverse professional backgrounds, utilizes a lean, remote-collaboration model to manage multiple projects simultaneously, bypassing traditional film crew structures.
The production group Can Diep, founded by 54-year-old Truong Han Sinh, transitioned from developing AI products for partners to creating its own intellectual property. The core team for the 'Song Dynasty' project consists of 10 members who handle all aspects of production, including scriptwriting, editing, and IP management.
To address the common AI challenge of character inconsistency, the team developed specific image datasets for each character. Production requires significant human intervention, with complex war sequences requiring between 10 and 20 takes to achieve the desired composition and camera movement.
The production team claims the project has been profitable, reporting over 1 million yuan in AI computing costs against tens of millions of yuan in commercial revenue. However, these figures are self-reported by the production company and have not been independently verified.
The project's workflow relies on a decentralized, remote structure, allowing the team to operate without the overhead of traditional film departments. This model is increasingly common in the Chinese market, where DataEye reports that over 220,000 AI-generated short films were published on Douyin in the first half of 2026 alone.
Why it matters
The success of this series illustrates a practical, scalable application of in the entertainment industry, demonstrating how small teams can produce high-volume content at a fraction of traditional costs. By replacing specialized departments—such as set designers, cinematographers, and traditional VFX artists—with AI-driven workflows, the production team has achieved profitability. This shift is part of a broader trend in the Chinese digital market, where the cost of producing AI-generated short films has dropped from 3,000 yuan per minute in 2025 to between 200 and 300 yuan per minute today, significantly lowering the barrier to entry for content creators.
The drastic reduction in production costs—from 3,000 yuan per minute in 2025 to 200-300 yuan per minute in 2026—signals a commoditization of digital video production. This allows for rapid iteration and the ability to release content on a weekly schedule.
The reliance on AI for visual effects and character generation allows small, non-traditional teams to compete with larger studios. This democratization of content creation is shifting the focus from high-budget production values to volume and algorithmic reach.
The report highlights a critical limitation: the reliance on human oversight for quality control and post-production. Despite the automation of image generation, the '' requirement remains essential for maintaining narrative coherence and professional standards.
The commercial success of this series serves as a case study for the potential profitability of AI-native content, though the lack of independent financial audits means the true scale of the industry's profitability remains an unknown.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
crm_get_transaction(id='4092').Which component of an AI application is the machine-learning model itself?
What to watch next
Industry observers should monitor the sustainability of this 'lean' production model as the market becomes increasingly saturated. While the report notes that 1,055 AI-generated works on Douyin reached over 100 million views in the first half of 2026, this represents less than 0.5% of the total volume of new AI content published on the platform. The long-term viability of these projects remains dependent on the ability of small teams to maintain visual consistency and narrative quality, as well as the evolving policies of major streaming platforms regarding AI-generated intellectual property.
Watch for potential regulatory shifts in China regarding the labeling and copyright of AI-generated content, which could impact the monetization strategies of these production groups.
Monitor whether the current trend of high-volume, low-cost AI content leads to market fatigue or if it forces traditional media companies to adopt similar AI-integrated workflows to remain competitive.
Observe the evolution of AI tools specifically designed for video consistency, as these will directly dictate the quality ceiling for small-team productions.
Note that while the report cites 6 billion views, the specific metrics for 'views' on these platforms are often proprietary and may not equate to unique audience engagement or long-term retention.