← Back to community benchmarks

Hermes-3-Llama-3.1-8B

M4 Max (32c) · 36 GB · 4bit · 2026-05-03
Performance
1k
tokens
726.4
PP tok/s
71.7
TG tok/s
1411
TTFT (ms)
5.0
Peak mem (GB)
Hardware
Chip M4 Max (32c)
Memory 36 GB
GPU Cores 32
Software
oMLX v0.3.8
macOS macOS 15.7.3
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 726.4 71.7 5.0 GB current
4k 718.1 64.9 5.5 GB view
Batching Results
Batch Size TG tok/s Speedup
71.7 1.00×
132.6 1.85×
146.8 2.05×