← Back to community benchmarks

Qwen3.5-122B-A10B

M3 Ultra (80c) · 512 GB · 4bit · 2026-03-20
Performance
4k
tokens
983.2
PP tok/s
50.3
TG tok/s
4166
TTFT (ms)
67.9
Peak mem (GB)
Hardware
Chip M3 Ultra (80c)
Memory 512 GB
GPU Cores 80
Software
oMLX v0.2.19
macOS macOS 15.7.2
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 916.3 53.4 66.4 GB view
4k 983.2 50.3 67.9 GB current
8k 962.4 47.3 68.5 GB view
Batching Results
Batch Size TG tok/s Speedup
1× 53.4 1.00×
8× 193.1 3.62×