Just now, the AI video generation project Everlyn, which is associated with @Sui , announced that Sui has invested $250 million in it. This is another significant investment by Sui to enhance its AI ecosystem! Everlyn's vision is to become the fastest video AI model in the world built on a decentralized infrastructure layer.
🔹 Project Introduction
Everlyn AI is an open-source foundational video model and multimodal ecosystem aimed at bringing an open dream factory to the world. Its flagship model, Openlyn, has redefined AI video capabilities in terms of video output duration, interactivity, generation latency, and hyper-realistic visual quality.
The Everworld platform revolutionarily allows users to interact with photo-realistic virtual selves through personalized interactive video agents, executing any task. Combined with the Lyn Protocol, which is an on-chain layer providing permanence, provenance, and ownership of AI videos: decentralized reasoning, agent ownership, and video fingerprinting ensure content is tamper-proof.
💡 From Meta's Make-a-Video to Tencent's foundational models, Everlyn has driven significant breakthroughs like Google VideoPoet and Open-SORA-Plan. Now, it transforms AI video from content into verifiable digital memories, supporting modular skills such as on-chain scheduling posts, pulling token prices, and triggering contracts.
👥 Team Members
Everlyn's standout feature is its luxurious team members. Most project members originally came from Meta, and given Sui's deep Meta background, this investment is understandable.
Dr. Sernam Lim (CEO): Led Meta AI for 6 years, has over 100 visual and integrity papers, and expanded a user detection system for hundreds of millions. From deep fakes to dream factories, he promotes the autonomy of video AI.
Dr. Harry Yang (CTO): Chief author of Meta Make-a-Video, with over 60 papers and 5K+ citations; built Reels AI, deep fake defenses, and autoregressive generation models; architect of Everlyn-1.
Dr. Serge Belongie: Formerly at Google, co-founder, led top AI research.
Dr. Philip Torr: Formerly at OpenAI, co-founder, focused on next-generation multimodal AI systems.
Other contributors: Dr. Qifeng Chen, Dr. Ming-Hsuan Yang, and other top scientists from HKUST, Peking University, Harvard, and Tencent.
Yann LeCun (Chief Scientist at Meta AI) also tweeted to congratulate the team: “Two former Meta colleagues sernamlim and leehomyc have built a cool video generation product at Everlyn.”
