← Back to community benchmarks

Qwen3.6-27B-FP8-source-oQ4-mtp

M4 Max (40c) · 128 GB · 4bit · 2026-05-07

Performance

4k

tokens

175.0

PP tok/s

21.0

TG tok/s

23410

TTFT (ms)

18.1

Peak mem (GB)

Hardware

Chip M4 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.3.9.dev1

macOS macOS 26.4.1

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	223.8	23.7	16.7 GB	view
4k	175.0	21.0	18.1 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	23.7	1.00×
2×	30.2	1.27×
4×	30.7	1.30×