← Back to community benchmarks

Qwen3.5-4B

M1 (8c) · 8 GB · 4bit · 2026-03-16
Performance
1k
tokens
137.5
PP tok/s
21.6
TG tok/s
7449
TTFT (ms)
3.9
Peak mem (GB)
Hardware
Chip M1 (8c)
Memory 8 GB
GPU Cores 8
Software
oMLX v0.2.13
macOS macOS 26.3.1
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 137.5 21.6 3.9 GB current
4k 143.6 19.8 4.2 GB view
Batching Results
Batch Size TG tok/s Speedup
21.6 1.00×
32.8 1.52×
34.1 1.58×