← Back to community benchmarks

Qwen3.5-4B

M2 (10c) · 8 GB · 4bit · 2026-03-11
Performance
1k
tokens
201.6
PP tok/s
32.0
TG tok/s
5079
TTFT (ms)
3.9
Peak mem (GB)
Hardware
Chip M2 (10c)
Memory 8 GB
GPU Cores 10
Software
oMLX v0.2.7
macOS macOS 26.3.1
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 201.6 32.0 3.9 GB current
4k 194.4 25.0 4.2 GB view
Batching Results
Batch Size TG tok/s Speedup
32.0 1.00×
44.5 1.39×