← Back to community benchmarks

Qwen3.5-27B

M4 Max (32c) · 36 GB · 6bit · 2026-03-11
Performance
4k
tokens
168.3
PP tok/s
14.3
TG tok/s
24343
TTFT (ms)
23.2
Peak mem (GB)
Hardware
Chip M4 Max (32c)
Memory 36 GB
GPU Cores 32
Software
oMLX v0.2.7
macOS macOS 26.3.1
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 194.6 15.5 22.9 GB view
4k 168.3 14.3 23.2 GB current
Batching Results
Batch Size TG tok/s Speedup
15.5 1.00×
22.1 1.43×
20.6 1.33×