Open Source · MarkTechPost ·
Liquid AI Releases LFM2.5-Encoder-230M and LFM2.5-Encoder-350M: Bidirectional Encoders That Stay Fast at 8K Context on CPU
Liquid AI released two open-weight bidirectional encoder models, LFM2.5-Encoder-230M and 350M, with 8,192-token context windows based on its LFM2 hybrid backbone. The 350M model ranked fourth across a 17-task evaluation, while the 230M completes an 8K-token CPU pass in about 28 seconds.