← Back to community benchmarks

Qwen3.5-27BMTP

M5 Max (40c) · 64 GB · 4bit · 2026-05-17

Performance

8k

tokens

697.8

PP tok/s

37.9

TG tok/s

11739

TTFT (ms)

18.6

Peak mem (GB)

Hardware

Chip M5 Max (40c)

Memory 64 GB

GPU Cores 40

Software

oMLX v0.3.9.dev2

macOS macOS 26.4.1

Context 8,192

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	782.7	48.4	16.1 GB	view
4k	777.9	44.4	17.6 GB	view
8k	697.8	37.9	18.6 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	48.4	1.00×
2×	48.4	1.00×
4×	57.1	1.18×