← Back to community benchmarks

qwen2.5-0.5B-sft

M4 Pro (16c) · 48 GB · unknown · 2026-03-12
Performance
4k
tokens
4,304
PP tok/s
158.3
TG tok/s
952
TTFT (ms)
1.7
Peak mem (GB)
Hardware
Chip M4 Pro (16c)
Memory 48 GB
GPU Cores 16
Software
oMLX v0.2.8
macOS macOS 15.1
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 2,895 185.0 1.3 GB view
4k 4,304 158.3 1.7 GB current
Batching Results
Batch Size TG tok/s Speedup
185.0 1.00×
159.0 0.86×
306.1 1.65×