← Back to community benchmarks

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

M3 Max (40c) · 128 GB · 4bit · 2026-04-01
Performance
1k
tokens
445.8
PP tok/s
45.9
TG tok/s
2297
TTFT (ms)
65.6
Peak mem (GB)
Hardware
Chip M3 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.0
macOS macOS 26.4
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 445.8 45.9 65.6 GB current
4k 481.2 42.0 67.0 GB view
8k 421.7 39.6 67.7 GB view
Batching Results
Batch Size TG tok/s Speedup
45.9 1.00×
62.2 1.36×
83.9 1.83×