← Back to community benchmarks

gemma-4-31b-it

M5 Max (40c) · 128 GB · bf16 · 2026-05-23
Performance
8k
tokens
688.8
PP tok/s
7.2
TG tok/s
11893
TTFT (ms)
61.8
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.9
macOS macOS 26.5
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 641.2 7.3 58.8 GB view
4k 755.5 7.3 61.3 GB view
8k 688.8 7.2 61.8 GB current
16k 609.1 7.1 63.1 GB view
32k 538.2 6.7 66.0 GB view
64k 393.1 5.9 72.0 GB view
Batching Results
Batch Size TG tok/s Speedup
7.3 1.00×
7.5 1.03×
15.2 2.08×
29.3 4.01×