Skip to content
llm-speed
Leaderboard/model/qwen2-5-14b-instruct-4bit

Qwen2.5-14B-Instruct-4bit

1 workload result across 1 hardware configuration.

Fastest known config

14.4 decode tok/s

on M3 Pro (18-core GPU) + 36GB unified via mlx (4bit) see full run

M3 Pro (18-core GPU) + 36GB unifiedM3 Pro (18-core GPU) + 36GB unified

WorkloadBackendQuantdecode tok/sprefill tok/sTTFTRun
chat-shortmlx@0.31.34bit14.41tok/s117.9tok/s1,111msr_ia73dzeue0b

Qwen2.5-14B-Instruct-4bit on hardware