Liquid AI releases a 230M model optimized for phones, Raspberry Pi, and robots

mpfect 17 points 1 comment July 01, 2026
www.liquid.ai · View on Hacker News

Discussion Highlights (1 comments)

potus_kushner

these LFM 2.5 models are crazy fast. the (biggest in series) 8B-A1B model produces 35-40 t/s on an aged 6-core CPU using llama.cpp. it's my go-to model for whenever i need fast local inference. it's also pretty good at toolcalling. would love to see more finetunes on HF, but it appears not many people discovered it yet.

Semantic search powered by Rivestack pgvector
14,015 stories · 131,331 chunks indexed