← Back to community benchmarks

Harmonic-Hermes-9B-oQ4

M4 Pro (16c) · 24 GB · 4bit · 2026-04-17
Performance
32k
tokens
285.2
PP tok/s
33.8
TG tok/s
114912
TTFT (ms)
9.2
Peak mem (GB)
Hardware
Chip M4 Pro (16c)
Memory 24 GB
GPU Cores 16
Software
oMLX v0.3.6
macOS macOS 26.4.1
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 318.2 49.6 5.9 GB view
4k 314.7 48.3 6.5 GB view
8k 322.0 43.1 6.9 GB view
16k 309.2 42.3 7.7 GB view
32k 285.2 33.8 9.2 GB current
Batching Results
Batch Size TG tok/s Speedup
49.6 1.00×
64.5 1.30×
68.7 1.39×
66.4 1.34×