← Back to community benchmarks

Qwen2.5-7B-Instruct

M1 Pro (14c) · 16 GB · 4bit · 2026-04-28
Performance
1k
tokens
237.2
PP tok/s
36.3
TG tok/s
4318
TTFT (ms)
4.8
Peak mem (GB)
Hardware
Chip M1 Pro (14c)
Memory 16 GB
GPU Cores 14
Software
oMLX v0.3.8.dev3
macOS macOS 26.3.1
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 237.2 36.3 4.8 GB current
4k 249.2 32.9 4.9 GB view
8k 225.4 28.0 5.1 GB view
Batching Results
Batch Size TG tok/s Speedup
36.3 1.00×
47.3 1.30×
50.2 1.38×