← Back to community benchmarks

Qwen3.5-VL-4B-8bitCRACK

M3 Pro (14c) · 36 GB · 8bit · 2026-03-10
Performance
4k
tokens
486.5
PP tok/s
26.6
TG tok/s
8419
TTFT (ms)
6.1
Peak mem (GB)
Hardware
Chip M3 Pro (14c)
Memory 36 GB
GPU Cores 14
Software
oMLX v0.2.7
macOS macOS 26.3.1
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 480.6 27.9 5.9 GB view
4k 486.5 26.6 6.1 GB current
Batching Results
Batch Size TG tok/s Speedup
27.9 1.00×
54.7 1.96×
80.7 2.89×