Build specialized multimodal models with LFM3

Liquid AI has released LFM3, a family of six open-weight multimodal models designed for local processing, tool use, and structured output. Built on a new tri-hybrid architecture, the 1.7B parameter models feature a 256K context window and are optimized to run on low-resource hardware. This release expands the options for developers looking to deploy specialized AI agents that operate efficiently without relying on cloud-based infrastructure.