Liquid AI Releases LFM3: Six Open-Weight Multimodal Models for On-Device Deployment

Liquid AI has unveiled LFM3, a family of six open-weight models ranging from 1.7 to 2 billion parameters, designed for specialized, on-device AI tasks across text, vision, audio, and combined modalities. The announcement on October 27 introduces a significant shift in model architecture and efficiency, positioning the LFM3 lineup as a leading option for enterprises seeking local AI deployment without reliance on cloud infrastructure.