iFlytek debuts digital human creator

Chinese artificial intelligence pioneer iFlytek has unveiled a platform which simplifies and cuts the time it takes to create a virtual, or digital human from hours of video capture to just a single photograph, enabling far more natural and complex interactions.

The launch of iFlytek's "ultra-realistic digital human real-time generation technology" opens new possibilities for wider business application, and underscores the accelerating shift from passive visual avatars to active, intelligent digital workforces, as generative AI reshapes the economics of virtual interaction, experts said.

Traditionally, producing a lifelike digital human required about two hours of human video recording and lengthy training cycles. The new system, demonstrated at the launch event, generates a digital human instantly from one still image, complete with natural facial expressions, gestures, and lip-sync.

More notably, the real-time avatar can perform intricate actions such as picking up objects, drinking, singing, and dancing, delivering smoother and more immersive human-machine interaction, the company said.

According to iFlytek, the achievement rests on several core technologies: iFlytek's Spark voice large language model powers fluent dialogue in general knowledge domains; curated vertical-industry knowledge bases reduce hallucination rates and provide precise professional answers; and customizable personas with long-term memory enable the digital human to maintain consistent character traits and accumulate interaction history over time.

Gao Jingwen, general manager of the virtual digital human business department at the iFlytek Open Platform, said current deployment scenarios span brand livestreaming, customer service, and unmanned retail stores. As large-model adoption lowers cost barriers, digital humans are transitioning from a premium tool for large enterprises to a standard asset for a broader range of businesses, iFlytek said.

The launch comes as China's digital human market enters a critical scaling phase. According to iiMedia Research, the core digital human market in China reached 48.06 billion yuan ($7.12 billion) in 2025 and is projected to hit 93.56 billion yuan by 2030.

Yet industry observers note a persistent "hot market, cold deployment" gap, as many digital humans remain script-following displays unable to sustain longform real-time interaction.

Competitors are racing to close that gap. For instance, Tencent's game-streaming arm unveiled a real-time multimodal digital human foundation model that also generates interactive avatars from a single photo, supporting 24/7 livestreaming, singing, dancing and interactive gaming, earlier this month.

Baidu has also debuted a digital human video podcast solution which the company said can cut production costs by 74 percent while boosting content expressiveness considerably.

As large language model adoption drives down costs, digital humans are transitioning from a premium tool for large enterprises to a standard asset for businesses of all sizes, said Pan Helin, a member of the expert committee on information and communication economy of the Ministry of Industry and Information Technology.

masi@chinadaily.com.cn