← Back to community benchmarks

Qwen3.5-122B-A10B9bit

M3 Ultra (80c) · 256 GB · 8bit · 2026-03-30
Performance
16k
tokens
872.8
PP tok/s
36.9
TG tok/s
18772
TTFT (ms)
132.8
Peak mem (GB)
Hardware
Chip M3 Ultra (80c)
Memory 256 GB
GPU Cores 80
Software
oMLX v0.2.24
macOS macOS 26.3.1
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 741.8 43.3 129.6 GB view
4k 913.8 42.0 131.0 GB view
8k 919.0 40.1 131.6 GB view
16k 872.8 36.9 132.8 GB current
32k 765.6 31.4 135.2 GB view
64k 582.7 24.1 140.1 GB view
128k 361.1 16.0 149.7 GB view