← Back to community benchmarks

Qwen3.5-27B-Claude-4.6-Opus-Distilled

M5 Max (32c) · 36 GB · 6bit · 2026-03-18

Performance

4k

tokens

660.4

PP tok/s

17.5

TG tok/s

6202

TTFT (ms)

23.6

Peak mem (GB)

Hardware

Chip M5 Max (32c)

Memory 36 GB

GPU Cores 32

Software

oMLX v0.2.19.dev1

macOS macOS 26.3.1

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	594.6	17.7	22.1 GB	view
4k	660.4	17.5	23.6 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	17.7	1.00×
2×	29.9	1.69×
4×	35.1	1.98×