← Back to community benchmarks

Laguna-S-2.1-NVFP4

M5 Max (40c) · 128 GB · 4bit · 2026-07-24
Performance
16k
tokens
244.5
PP tok/s
7.3
TG tok/s
67006
TTFT (ms)
68.8
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.5.3
macOS macOS 27.0
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 226.5 5.8 67.7 GB view
4k 254.4 5.6 67.9 GB view
8k 210.7 6.3 68.3 GB view
16k 244.5 7.3 68.8 GB current
32k 235.5 6.4 70.0 GB view
64k 229.7 4.6 72.3 GB view
Batching Results
Batch Size TG tok/s Speedup
5.8 1.00×
9.7 1.67×
11.9 2.05×