← Back to community benchmarks

Qwen2.5-7B-Instruct

M4 (10c) · 24 GB · 4bit · 2026-03-09
Performance
1k
tokens
233.0
PP tok/s
22.1
TG tok/s
4395
TTFT (ms)
4.5
Peak mem (GB)
Hardware
Chip M4 (10c)
Memory 24 GB
GPU Cores 10
Software
oMLX v0.2.6
macOS macOS 26.3.1
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 233.0 22.1 4.5 GB current
4k 227.6 20.7 4.8 GB view
Batching Results
Batch Size TG tok/s Speedup
22.1 1.00×
43.6 1.97×
58.1 2.63×