← Back to community benchmarks

Qwen3.5-122B-A10BQ3-VL

M2 Max (38c) · 96 GB · 3bit · 2026-03-15
Performance
4k
tokens
334.4
PP tok/s
11.1
TG tok/s
12251
TTFT (ms)
54.5
Peak mem (GB)
Hardware
Chip M2 Max (38c)
Memory 96 GB
GPU Cores 38
Software
oMLX v0.2.13
macOS macOS 15.7.4
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 304.3 11.6 53.8 GB view
4k 334.4 11.1 54.5 GB current
Batching Results
Batch Size TG tok/s Speedup
11.6 1.00×
19.9 1.72×
39.2 3.38×