Liquid AI: Bringing LLMs to the Edge
“Why run your AI in the cloud when it can live and learn on your phone?”
Liquid AI has built something radical: Liquid Foundation Models.
🧠 Inspired by biological brains
📱 Trained and inferenced on-device — sub-100ms latency
🧩 1/6 the size of LLaMA 70B models — same performance
🔒 Built for privacy, personalization, and ultra-low compute
Backed by MIT, led by robotics pioneers, and now partnered with AMD.
💡 Building edge-native or privacy-first AI? Let’s explore together.