← Back to community benchmarks

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

M3 Ultra (60c) · 96 GB · 3bit · 2026-04-05
Performance
16k
tokens
2,260
PP tok/s
20.1
TG tok/s
7249
TTFT (ms)
62.4
Peak mem (GB)
Hardware
Chip M3 Ultra (60c)
Memory 96 GB
GPU Cores 60
Software
oMLX v0.3.2
macOS macOS 26.4
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
4k 734.6 51.4 60.3 GB view
16k 2,260 20.1 62.4 GB current
64k 1,896 18.9 65.7 GB view