How AI Is Turning Prompts into Chinese Micro-Dramas

Fans of China’s fast-growing micro-dramas got a glimpse behind the scenes at a recent trade expo, where artificial-intelligence tools were shown turning written prompts into characters, scenes and moving images.

From text prompts to animated clips

At the 23rd China-ASEAN Expo in Nanning, a staff member from Nova AI walked through the platform’s workflow on September 18. A user begins by entering text instructions. The system then creates an image and, after additional selections and prompts, converts it into video.

In one demonstration, the platform first produced a fantasy character against a dark background. The operator then used editing options to animate the still image. Within a short period, a brief sequence appeared showing cloaked figures moving through a mountain landscape.

The example showed how a single interface can carry a rough idea from written description to visual and animated output. Such tools are helping Chinese producers reduce some production time and costs for micro-dramas, advertisements and other online video content.

A format built for mobile viewing

Micro-dramas are different from traditional television series. They are mainly designed for smartphone screens, with episodes lasting one or two minutes and fast-moving stories that encourage viewers to keep watching the next installment.

In Cambodia, many of these series reach audiences through Khmer subtitles, translated narration or synthetic voices shared on social-media and video platforms. Numerous Khmer-language pages and channels carry Chinese series, either as individual short episodes or edited into longer compilations.

It is not always clear whether those translations are produced entirely by AI, checked by human translators or made through a mix of both. Some programmes are traditionally filmed dramas translated with computer assistance. Others use AI for parts of the process, such as writing, image creation, editing, dubbing or visual effects, while a smaller number rely completely on AI-generated imagery and video.

Nova AI and the wider production chain

Nova AI is one of the Chinese companies targeting this expanding production space. Zhou Peixun, vice-president of Nova AI, said the platform brings together domestic and overseas AI models and lets users choose tools for different content tasks.

He described scriptwriters, directors and companies producing micro-dramas as likely users. Zhou also told ASEAN journalists that the technology is not limited to drama production: it can also support e-commerce, advertising and other online sectors.

The system does not create a full drama with a single button press. Instead, it offers tools that can assist in areas such as script development, character design, image generation, storyboarding and video production. Creators still need to provide instructions, choose images, revise material and decide how sequences fit together into a complete story.

During the demonstration, the operator could enter a written description, select a visual-generation model and specify the desired format. The resulting image then became the basis for a short animated clip.

Stardust and the scale of the market

Presenters also highlighted Stardust, an associated micro-drama application that gives users access to more than 12,000 short dramas from China and ASEAN countries. They said the service supports 14 languages and that around 30 new dramas are produced each month.

They cited an AI-assisted production cost of between 200 and 300 yuan, or about $30 to $45, per minute, depending on the model used. That figure refers to video generation through the system, rather than the full cost of producing, editing and distributing a completed drama. One presenter claimed that micro-dramas generate more than $50 million in box office per month.

The Nova AI demonstration formed part of a broader CAEXPO focus on AI cooperation between China and ASEAN. Presenters said a China-ASEAN AI application cooperation centre in Guangxi had trained more than 800 students from ASEAN countries and recorded exchanges involving over 1,000 people. The centre has also sent more than 30 groups of specialists to ASEAN countries and supported over 30 cross-border AI projects covering smart agriculture, manufacturing, city management, cultural tourism and e-commerce.

What it could mean for Cambodia

For Cambodia, the technology carries two possible effects. It may deliver a larger volume of translated Chinese entertainment to local viewers. At the same time, it could give Cambodian writers, filmmakers, online sellers and advertising companies access to cheaper content-production tools.

Machine translation and synthetic voices can make dramas available quickly, but the quality remains uneven. Literal or inaccurate translations may change the meaning of dialogue, weaken emotional scenes or misrepresent cultural references. Human review is still important, especially for Khmer, where context, forms of address and relationships between speakers shape how dialogue should be rendered.

The growth of AI also raises questions about copyright and creative ownership. It is not always clear who owns a character, script, image or video produced using several AI models, particularly when those systems may have been trained on existing creative works. There are also concerns that automation could affect writers, translators, actors, voice performers, editors and other film-production professionals.

Whether Cambodian creators become production partners or mainly consumers of imported content remains unsettled. The tools may make production faster and cheaper, but their lasting impact will depend on translation quality, the protection of personal and creative data, and whether Cambodian storytellers can use the same technology to create and distribute their own stories.