← Back to community benchmarks

Qwen3.5-27B

M4 Max (40c) · 128 GB · 4bit · 2026-04-20
Performance
128k
tokens
362.3
PP tok/s
13.2
TG tok/s
361787
TTFT (ms)
53.8
Peak mem (GB)
Hardware
Chip M4 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.5.dev1
macOS macOS 26.3.1
Context 131,072
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 241.8 29.0 24.7 GB view
4k 250.0 27.9 26.1 GB view
8k 248.1 27.0 26.8 GB view
16k 480.5 22.9 28.5 GB view
32k 464.5 22.4 29.7 GB view
64k 428.4 21.1 31.6 GB view
128k 362.3 13.2 53.8 GB current
195k 307.9 18.0 70.4 GB view
Batching Results
Batch Size TG tok/s Speedup
29.0 1.00×
54.4 1.88×
62.8 2.17×