← Back to community benchmarks

Qwen3.5-VL-122B-A10B-4bitCRACK

M5 Max (40c) · 128 GB · 4bit · 2026-03-26
Performance
4k
tokens
1,596
PP tok/s
48.1
TG tok/s
2566
TTFT (ms)
67.9
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.2.21
macOS macOS 26.4
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,258 50.2 66.4 GB view
4k 1,596 48.1 67.9 GB current
Batching Results
Batch Size TG tok/s Speedup
50.2 1.00×
83.0 1.65×
114.1 2.27×