← Back to community benchmarks

Qwen3.5-27B7bit

M1 Ultra (48c) · 64 GB · 6bit · 2026-03-12

Performance

8k

tokens

129.7

PP tok/s

16.9

TG tok/s

63180

TTFT (ms)

24.4

Peak mem (GB)

Hardware

Chip M1 Ultra (48c)

Memory 64 GB

GPU Cores 48

Software

oMLX v0.2.9

macOS macOS 26.3.1

Context 8,192

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	125.9	18.2	23.6 GB	view
4k	130.0	17.7	23.9 GB	view
8k	129.7	16.9	24.4 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	18.2	1.00×
2×	22.5	1.24×
4×	24.9	1.37×