← Back to community benchmarks

gemma-4-31b-it

M5 Max (40c) · 128 GB · bf16 · 2026-05-25
Performance
16k
tokens
647.3
PP tok/s
7.2
TG tok/s
25312
TTFT (ms)
61.6
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.9
macOS macOS 26.5
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 641.2 7.4 58.8 GB view
4k 788.0 7.4 60.0 GB view
8k 725.4 7.3 60.5 GB view
16k 647.3 7.2 61.6 GB current
32k 510.1 7.1 63.9 GB view
64k 405.4 6.4 68.7 GB view
128k 275.9 5.2 78.2 GB view
Batching Results
Batch Size TG tok/s Speedup
7.4 1.00×
9.0 1.22×
16.3 2.20×