[Important] AI Video Generation Indistinguishable from Real People from a Single Photo! "DigiSelf" Service Launch Enables Interaction with AI Avatars through Ultra-Fast 0.01-Second Generation Technology

EmbodyMe Inc. (Headquarters: Tokyo, CEO: Kazuya Kudo), a developer of foundational technology for video generation AI, will launch its AI avatar video generation service "DigiSelf" on March 30, 2026. (URL: https://digiself.tech/)

DigiSelf is a service that can generate AI avatar videos of indistinguishable quality from real people, based on a single photo and text. It generates in real-time, enabling interactive communication with AI avatars.

Prior to its general release, a limited pre-release was conducted for select companies, receiving positive feedback from dozens of participating businesses. Notable use cases include the appearance of AI Tamura Jun on the radio program "Tamura Jun's News CLUB" on Cultural Broadcasting and AI Governor Yasutomo Suzuki's appearance at TECH BEAT Shizuoka.

Service Video:

Real-time Demo Video:

EmbodyMe's Journey So Far Since its establishment in 2016, EmbodyMe Inc. has been driving research and development of foundational technologies for generative AI in the video field. Our video generation AI technology boasts world-leading competitive advantage in real-time processing performance, with a track record of awards such as the NVIDIA Inception Award and Microsoft Innovation Award, as well as multiple international patents. Products leveraging this high technological capability are widely deployed both domestically and internationally, receiving high acclaim in the global market with 95% of users being from overseas. Notably, we achieved first place on "Product Hunt," a standard platform for product launches abroad. ■ DigiSelf Overview DigiSelf is a service capable of generating AI avatar video and audio of indistinguishable quality from real people. Based on a single photo or a one-minute video, and by inputting text...