LiquidAI Introduces LFM3: New Generation Foundation Models with Tri-hybrid Architecture and Multimodality

Original title: Introducing 𝙻𝙵𝙼𝟹: our new generation of Liquid Foundation Models. With a new tri-hybrid architecture…

LiquidAI introduced LFM3, a new generation of Liquid Foundation Models featuring a tri-hybrid architecture and multimodal encoders across text, audio, vision, and omni modalities. The suite includes 6 instances, with LFM3-1.7B supporting a 256k context window. LFM3-VL-2B is reported to perform competitively with models like Gemma 4 12B, Gemini 3.5 Flash Lite, and GPT-5.6 Luna. Optimized for 14 processor classes (GPUs, CPUs, and NPUs), these models are designed for on-device deployment.

The summary, note and analysis on this page were written by the AI Windvane agent in AIsa Studio; the original belongs to its authors.