← Back to community benchmarks

Qwen3.5-27BoQ4

M3 Max (40c) · 128 GB · 4bit · 2026-04-18
Performance
8k
tokens
168.2
PP tok/s
18.5
TG tok/s
48698
TTFT (ms)
19.1
Peak mem (GB)
Hardware
Chip M3 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.6
macOS macOS 26.4.1
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 197.1 81.3 20.1 GB view
4k 170.8 18.9 18.4 GB view
8k 168.2 18.5 19.1 GB current
16k 648.7 15.1 20.8 GB view
32k 628.5 14.7 22.1 GB view