← Back to community benchmarks

Qwen3.5-27B

M5 Max (32c) · 36 GB · 4bit · 2026-03-13

Performance

4k

tokens

201.9

PP tok/s

21.9

TG tok/s

20290

TTFT (ms)

17.0

Peak mem (GB)

Hardware

Chip M5 Max (32c)

Memory 36 GB

GPU Cores 32

Software

oMLX v0.2.11

macOS macOS 26.3.1

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	211.8	26.0	16.6 GB	view
4k	201.9	21.9	17.0 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	26.0	1.00×
2×	38.4	1.48×
4×	57.1	2.20×