← Back to community benchmarks

Qwen3.5-4B

M4 (10c) · 16 GB · 4bit · 2026-05-10
Performance
1k
tokens
386.6
PP tok/s
38.7
TG tok/s
2649
TTFT (ms)
3.3
Peak mem (GB)
Hardware
Chip M4 (10c)
Memory 16 GB
GPU Cores 10
Software
oMLX v0.3.9.dev1
macOS macOS 26.2
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 386.6 38.7 3.3 GB current
8k 390.5 34.9 4.5 GB view
16k 368.4 31.5 5.3 GB view
Batching Results
Batch Size TG tok/s Speedup
38.7 1.00×
72.6 1.88×
90.3 2.33×
180.8 4.67×