Vivix announces it will launch real-time interactive models W1 and A1 in July

Vivix founder Liu Yu publicly explained for the first time that the company's direction is next-generation real-time interactive models rather than video generation tools, and announced that in July it will release the foundation model W1, which integrates understanding, interaction, and generation, and A1, which is aimed at character interaction. According to his introduction, W1's real-time generation cost is already close to YouTube's CDN bandwidth cost of about $0.2 per hour, the first-frame generation time is about 300 to 600 milliseconds, and the full-chain first response time is about 1 second. Liu Yu said that over 19 months the company redid data collection, the data pipeline, and infra, has finely filtered tens of millions of hours of multimodal data, and plans to complete training of W2, with a data volume two orders of magnitude higher, in Q4. There is currently no information on progress after the release.

AI-generated from public reports

1. Oct 26· 1 reports

   Vivix announces that real-time interactive models W1 and A1 will be launched in July

   Source:36Kr · AI