← Back to community benchmarks

Qwen3.5-27BMTP

M3 Ultra (60c) · 96 GB · 8bit · 2026-05-14

Performance

1k

tokens

294.0

PP tok/s

31.9

TG tok/s

3483

TTFT (ms)

28.8

Peak mem (GB)

Hardware

Chip M3 Ultra (60c)

Memory 96 GB

GPU Cores 60

Software

oMLX v0.3.9.dev2

macOS macOS 26.3.1

Context 1,024

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	294.0	31.9	28.8 GB	current
4k	316.7	30.9	30.3 GB	view

Batching Results

Batch Size	TG tok/s	Speedup
1×	31.9	1.00×
2×	39.3	1.23×
4×	50.1	1.57×