Faceu Technology
- Chinese name
- 剪映
- Pinyin
- Registered name
- 深圳市脸萌科技有限公司
- Unified Social Credit Code
9144030008867405X2registered in Guangdong 广东- Website
- 7268064897
Shenzhen Faceu Technology (深圳市脸萌科技有限公司) is the ByteDance company behind Jianying (剪映), whose international counterpart CapCut is provided by Bytedance Pte. Ltd. in Singapore, together with the imaging apps Xingtu (醒图), Qingyan (轻颜相机) and Faceu (激萌), and the AI creation platform Jimeng (即梦AI). Incorporated in January 2014 in Shenzhen’s Nanshan district and wholly held through ByteDance’s Beijing Xingyun Chuangji Technology (北京星云创迹科技), the company grew from the viral Faceu avatar app into the group’s tool-product incubator; Jianying (剪映)’s privacy policy names it as the product’s owner and operator.
Digital humans are a headline line of the editor: Jianying Professional (剪映专业版) 4.6 introduced the digital-human creation tool, with avatar presenters selected from a library or customized, one-sentence voice cloning, and lip-synced spoken-presenter videos generated from a script — the workflow that has made presenter-fronted video production a mass-market feature, with the official site listing digital human (数字人) among its AI capabilities alongside smart packaging and marketing video generation. The registered legal representative is Shi Beini (施蓓妮), with encyclopedia sources carrying Li Fei (李飞) after a December 2025 management change; registered capital is RMB 10 million, fully paid.
Jimeng AI (即梦AI), known internationally as Dreamina, is ByteDance’s generative AI creation platform, built by the Jianying team; its user agreement names the company as owner and operator of the Jimeng app and of jimeng.jianying.com. The service entered closed beta in March 2024 under the original name Jianying Dreamina and has since grown into a multimodal creation suite covering text-to-image, image-to-video, voice generation, lip-sync, action imitation, and a dedicated digital human module, accessible through a consumer web and mobile app and through an enterprise API distributed via ByteDance’s Volcano Engine (火山引擎) cloud platform. The digital human stack is organized around two principal modes. The Quick Mode (快速模式) takes a single portrait image plus an audio clip and returns a synchronized talking-head video, while the Master Mode (大师模式), launched on March 7, 2025, is driven by ByteDance’s proprietary OmniHuman-1 model and accepts portrait, half-body, or full-body inputs, including non-photorealistic source images such as anime characters and 3D cartoons, producing lifelike video in which the figure speaks, sings, plays instruments, and moves naturally in accordance with an input audio track. A complementary Action Imitation (动作模仿) feature, also released in March 2025, transfers motion captured in a reference video onto a still image of a person. The underlying lip-sync engine was later upgraded to a 1.5 release, and a successor generation model, OmniHuman 1.5, further expanded controllability over expression, gesture, and whole-body kinematics. ByteDance positions the digital human stack as a low-cost replacement for traditional video shoots in scenarios including character design, presentation, online education, livestream commerce, and short-form social video, and offers a template gallery with more than one hundred avatar styles, including a TikTok-oriented variant aimed at cross-border marketing. In line with the Cyberspace Administration of China’s regulatory framework for synthetic media, the platform applies a content review mechanism and labels all digital human outputs with an AI-generated watermark. By the first quarter of 2026, Jimeng AI had reached approximately 13.52 million monthly active users and 5.59 million downloads, placing it among the most widely used domestic Chinese platforms for generative digital human video, and in December 2025 it launched a micro-short drama initiative, Dimensional Folding (次元折叠), built around its digital human and video generation pipeline.