← Back to community benchmarks

Qwen3-14B

M5 (10c) · 24 GB · 6bit · 2026-04-27
Performance
8k
tokens
396.3
PP tok/s
8.6
TG tok/s
20672
TTFT (ms)
12.5
Peak mem (GB)
Hardware
Chip M5 (10c)
Memory 24 GB
GPU Cores 10
Software
oMLX v0.3.7
macOS macOS 26.4.1
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 394.6 10.0 11.3 GB view
4k 423.0 9.7 11.9 GB view
8k 396.3 8.6 12.5 GB current
Batching Results
Batch Size TG tok/s Speedup
10.0 1.00×
17.1 1.71×
20.7 2.07×