Vivix has launched its first real-time interactive multimodal model, A1, designed to enable continuous video communication with virtual characters. The model allows characters to exhibit physical presence, interact with their environment, and respond dynamically to user input, creating a sense of real-time connection. To achieve this, Vivix developed a unified streaming architecture that integrates multimodal references, real-time interaction, and continuous video generation, while also introducing W1 to allow users to influence the narrative. The system boasts impressive efficiency, with a nearly 30B parameter model capable of running on consumer GPUs and achieving low latency through techniques like Multidimensional Joint Distillation (MJD) and a decoupled infrastructure. AI
IMPACT Enables highly interactive and embodied virtual characters, potentially transforming real-time communication and entertainment.
RANK_REASON Frontier-lab model release with system card [lever_c_demoted from frontier_release: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
