← Back to community benchmarks

Qwen3.5-122B-A10B-oQ3.5

M4 Max (40c) · 128 GB · 3bit · 2026-04-30
Performance
16k
tokens
485.3
PP tok/s
41.4
TG tok/s
33758
TTFT (ms)
61.2
Peak mem (GB)
Hardware
Chip M4 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.8rc1
macOS macOS 26.4.1
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 509.0 51.8 58.1 GB view
4k 573.5 49.6 59.4 GB view
16k 485.3 41.4 61.2 GB current
64k 321.9 24.9 68.5 GB view