← Back to community benchmarks

Qwen3.5-4B

M5 (8c) · 16 GB · 4bit · 2026-04-16
Performance
1k
tokens
913.7
PP tok/s
47.8
TG tok/s
1121
TTFT (ms)
3.3
Peak mem (GB)
Hardware
Chip M5 (8c)
Memory 16 GB
GPU Cores 8
Software
oMLX v0.3.5
macOS macOS 26.3.2
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 913.7 47.8 3.3 GB current
4k 1,004 45.8 3.9 GB view
Batching Results
Batch Size TG tok/s Speedup
47.8 1.00×
87.8 1.84×
101.4 2.12×