← Back to community benchmarks

Qwen3-4B-Instruct-25075bit

M3 Pro (18c) · 36 GB · 5bit · 2026-03-10
Performance
4k
tokens
529.7
PP tok/s
33.5
TG tok/s
7733
TTFT (ms)
3.9
Peak mem (GB)
Hardware
Chip M3 Pro (18c)
Memory 36 GB
GPU Cores 18
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 575.7 38.0 3.4 GB view
4k 529.7 33.5 3.9 GB current
Batching Results
Batch Size TG tok/s Speedup
1× 38.0 1.00×
2× 74.4 1.96×
4× 84.3 2.22×