Skip to content
llm-speed
Leaderboard/model/starcoder2-15b-4bit

starcoder2-15b-4bit

2 workload results across 2 hardware configurations.

Fastest known config

64.2 decode tok/s

on M3 Ultra (60-core GPU) + 96GB unified via mlx see full run

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

WorkloadBackendQuantdecode tok/sprefill tok/sTTFTRun
chat-shortmlx@0.31.3no datano datano datar_w7t2hm1p-9f

M3 Ultra (60-core GPU) + 96GB unifiedM3 Ultra (60-core GPU) + 96GB unified

WorkloadBackendQuantdecode tok/sprefill tok/sTTFTRun
chat-shortmlx@0.31.364.16tok/s220.5tok/s490msr_wsxml_39dh_

starcoder2-15b-4bit on hardware