← Back to community benchmarks

Qwen3.6-27B-UD

M5 Max (40c) · 48 GB · 4bit · 2026-06-04

Performance

4k

tokens

824.8

PP tok/s

20.6

TG tok/s

4966

TTFT (ms)

25.5

Peak mem (GB)

Hardware

Chip M5 Max (40c)

Memory 48 GB

GPU Cores 40

Software

oMLX v0.4.0rc2

macOS macOS 26.5

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	704.5	20.8	24.6 GB	view
4k	824.8	20.6	25.5 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	20.8	1.00×
2×	23.7	1.14×
4×	39.9	1.92×