Liquid AI Debuts LFM2.5-24B-A2B: A Fast 2B-Active MoE Model for On-Device Agents
Liquid AI has launched LFM2.5-24B-A2B, an open-weight Mixture-of-Experts reasoning model featuring 24 billion total parameters with only 2 billion active. Designed for local on-device agents, the model achieves strong benchmark scores and rapid decoding speeds on modern hardware, signaling a major step forward for efficient edge computing and local AI deployment.
Source: MarkTechPost