Top 50 Virtual Digital Human Companies in 2021 (2021虚拟数字人企业排名TOP50) is a Chinese Top-50 ranking list released by Internet Weekly that compares 50 participating companies in China’s “virtual digital human” sector across multiple forms of virtual beings, including virtual anchors/hosts, virtual employees, virtual idols, and ultra-realistic digital humans, using representative projects as evidence of capability and industry impact; the published table presents a numbered list with short exemplars for leading entries (for example, Baidu at No. 1, Alibaba Group at No. 2, and Tianyi Henian at No. 3) and is commonly cited in subsequent coverage as a snapshot of overall “comprehensive strength” in the domestic digital-human industry for 2021.
The piece argues that China’s virtual digital human sector has shifted from a largely conceptual “metaverse” talking point into industrial-scale deployment, driven by the global attention on Web3.0, major tech firms rolling out metaverse-related products, and viral consumer-facing examples that pulled the topic into mainstream discussion. It points to the virtual beauty influencer Liu Yexi as a 2021 catalyst for mass awareness, and to McDonald's selecting Xijiajia as a virtual endorser as a fresh trigger for industry and public debate, framing these events as signs that virtual beings are moving beyond niche fandom into commercial and institutional use.
It then centers on a ranking release by Internet Weekly titled “2021虚拟数字人企业排名TOP50,” described as a comparative list of 50 companies across multiple virtual digital human forms, including virtual hosts, virtual employees, virtual idols, and ultra-realistic digital humans. The article states that Baidu ranked first overall on “comprehensive strength,” citing representative deployments such as virtual broadcasting work with China Central Television and a Winter Olympics sign-language digital human host; Alibaba Group ranked second; and Tianyi Henian ranked third, identified as the producer behind the virtual singer Luo Tianyi. The ranking is presented as an industry snapshot that weights factors such as virtual IP influence and the breadth of participating firms’ core businesses.
To show how the category evolved, the article traces a loose lineage from early mass-known virtual performers to today’s broader ecosystem. It references Hatsune Miku as an early widely recognized example in the late 2000s, then notes the continued rise of performance-oriented and celebrity-style digital humans, including A-SOUL created by Yuehua Entertainment with creative participation from Fang Wenshan and Xu Song, and the ultra-realistic digital human AYAYI, described as going viral from a single social-media image and later collaborating with brands such as Louis Vuitton and Guerlain. Alongside these “performance” cases, it emphasizes that service-oriented deployments have become commonplace, positioning virtual digital humans as practical interfaces for broadcasting, customer service, and enterprise front-desk roles rather than only as entertainment properties.
The article uses the Beijing Winter Olympics period to illustrate service impact at scale, describing a sign-language virtual anchor built by Baidu AI Cloud that provided continuous sign-language coverage for hearing-impaired audiences, and a widely shared on-air interaction framed as a “PK” with Zhu Guangquan. It adds other institutional examples, including China Mobile presenting a digital “avatar” of Gu Ailing and enterprises adopting virtual employees such as a customer-service avatar for China Unicom and a banking avatar for Shanghai Pudong Development Bank. As supporting context, it cites a report from Communication University of China that groups China’s virtual digital humans into identity-type (real-person digital doubles), service-type (virtual employees), and performance-type (virtual idols), and claims the industry is entering a surge phase aided by policy support, a maturing supply chain, audience growth, and capital interest.
A major through-line is cost and production efficiency. The article quotes Li Shiyan describing three obstacles to large-scale rollout: fragmented production and optimization workflows across the supply chain, a weak connection between performance-oriented and service-oriented capabilities (with one lacking business competence and the other lacking “persona”), and persistently high costs for high-frequency, high-mobility needs that ultimately trace back to low production efficiency. As Baidu’s answer, it highlights the Baidu Intelligent Cloud Xiling platform, described as combining digital human production, content creation, and business configuration services, supported by multiple AI engines for avatar driving, dialogue, speech interaction, and recommendation, and positioned for use across broadcasting, entertainment, finance, government services, telecom, and retail.
The conclusion ties commercialization to “cost reduction and efficiency gains” in both avatar creation and content generation. It cites market sizing figures attributed to iMedia Research (艾媒咨询), claiming the virtual idol segment grew strongly in 2021 and is projected to expand further in 2022, and it references Wu Tian discussing a goal of compressing production cycles toward hours and using AI to generate content at scale, a framing aligned with AIGC-style workflows showcased through Xijiajia. The article’s headline claim is that, as platformization and AI automation improve, production timelines could shrink dramatically—2D avatars from about a week to minutes and ultra-realistic 3D avatars from months to roughly one to two weeks—while costs could fall from “million-level” RMB budgets to “ten-thousand-level” RMB budgets within about two years, culminating in a popularly phrased goal of broad, everyday access to personally usable digital humans.