Tech Jun 8, 2026 17 min LFM2.5 1.2B JP on M1 Max 64GB: 208 tok/s decode, JSON OK, name hallucinated Tested LFM2.5-1.2B-JP-202606 on M1 Max 64GB. llama.cpp Q4_K_M: 208 tok/s decode, JSON intact, model name hallucinated (LFM→FDM). Q8_0: 157 tok/s, no hallucination. Tool calls broken via GGUF. AI LLM Local LLM MLX Ollama Apple Silicon Edge AI Experiment Japanese LLM
