← Back to community benchmarks

Qwen3.5-122B-A10B

M5 Max (40c) · 128 GB · 4bit · 2026-04-01
Performance
128k
tokens
601.3
PP tok/s
20.5
TG tok/s
217972
TTFT (ms)
85.7
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.0
macOS macOS 26.4
Context 131,072
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,052 66.5 65.6 GB view
4k 1,516 63.6 67.0 GB view
8k 1,545 60.2 67.7 GB view
16k 1,429 52.3 68.9 GB view
32k 1,173 42.7 71.2 GB view
64k 900.4 31.4 76.1 GB view
128k 601.3 20.5 85.7 GB current
195k 428.2 14.6 95.9 GB view
Batching Results
Batch Size TG tok/s Speedup
66.5 1.00×
101.3 1.52×
134.8 2.03×