Liquid AI launches LFM3, a multimodal model family built for specialization
Discovery source: Liquid AI on Hugging Face · Source date:
Liquid AI says LFM3 is its new generation of Liquid Foundation Models, designed as a starting point for specialized intelligence across text, vision, audio, omni, and decision use cases. The company says the family uses one shared backbone and stays in the 2B-parameter range, so teams can fine-tune a chosen variant on a single GPU and deploy it on device or in their application environment. Liquid AI also claims LFM3-1.7B outperforms LFM2.5-2.6B with 35% fewer parameters, but those performance results are the company’s own measurements. The announcement adds that the family includes six models, including first omni support and video input for vision, plus a 256K context window for agentic workloads.
Check the sources
This is a NextGen-AI briefing, not a reproduction of the discovery source. A newsletter may cover several developments; consult the evidence below for this briefing’s specific subject. The source date refers to the discovery entry, not necessarily the announcement date.