← Back to community benchmarks

Qwen3.5-122B-A10B

M3 Ultra (80c) · 512 GB · 8bit · 2026-03-26
Performance
16k
tokens
879.4
PP tok/s
28.9
TG tok/s
18631
TTFT (ms)
126.5
Peak mem (GB)
Hardware
Chip M3 Ultra (80c)
Memory 512 GB
GPU Cores 80
Software
oMLX v0.2.21
macOS macOS 26.3.1
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 855.6 33.0 123.3 GB view
4k 952.4 31.9 124.8 GB view
8k 935.5 30.6 125.4 GB view
16k 879.4 28.9 126.5 GB current
32k 764.9 26.2 128.9 GB view
64k 579.6 22.0 133.8 GB view
128k 378.6 16.7 143.6 GB view
195k 276.5 13.3 153.6 GB view