LFM2.5-24B-A2B is a 24B-parameter Mixture-of-Experts model with 2B active parameters per token. It outperforms Qwen3.6-35B-A3B on agentic benchmarks like AutomationBench and achieves 184 tokens/s on Apple M5 Max hardware.
HOW THIS AFFECTS YOU
●
builderYou can deploy high-speed reasoning agents on edge devices using the open-weight models.
●
founderThis provides a competitive path for building private, low-latency agentic applications without massive cloud costs.