Liquid AI releases LFM2.5-24B-A2B, a sparse model built for fast on-device agents
gmays · hn · 2026-10-15
Liquid AI announced LFM2.5-24B-A2B, a 24B-total-parameter (≈2B active) sparse model optimized for running agents quickly on local devices, with a focus on on-device inference speed and agentic workflows.
More from Infra
- SK hynix up ~8x in a year as AI industry begs for cheap memory — AIFlow_ML · 2026-10-15
- Disaggregated Inference Is an Operations Story Too, Not Just Tokens Per Second — ai · 2026-10-15
- Micro Data Center Build Under Fire for Workers Standing on Forklift Tines — oilmutt · 2026-10-15
- Nvidia's Ian Buck on Why GPUs Keep Winning: Blackwell Perf Up 4x Since Launch — firstadopter · 2026-10-15
- AI-rewritten libtiff closes gap to ~10x of C version, fueling hopes of retiring buggy C libs — jd_pressman · 2026-10-15
- PyTorch Day Japan 2026 agenda live: Tokyo event Dec 10, early-bird saves ¥3,000 — PyTorch · 2026-10-15