← Back to community benchmarks

Qwen2.5-Coder-14B-Instruct

M3 Max (30c) · 36 GB · 4bit · 2026-04-15

Performance

4k

tokens

315.9

PP tok/s

27.2

TG tok/s

12967

TTFT (ms)

9.2

Peak mem (GB)

Hardware

Chip M3 Max (30c)

Memory 36 GB

GPU Cores 30

Software

oMLX v0.3.5.dev1

macOS macOS 26.3.1

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	320.7	30.8	8.5 GB	view
4k	315.9	27.2	9.2 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	30.8	1.00×
2×	51.5	1.67×
4×	63.8	2.07×