← Back to community benchmarks

gemma-4-31B-it

M5 Max (40c) · 128 GB · 8bit · 2026-06-01
Performance
16k
tokens
309.4
PP tok/s
7.2
TG tok/s
52946
TTFT (ms)
36.1
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.12
macOS macOS 26.4.1
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 317.6 9.9 31.8 GB view
4k 366.2 9.4 34.5 GB view
8k 349.2 8.6 35.2 GB view
16k 309.4 7.2 36.1 GB current
32k 247.2 5.7 39.1 GB view
64k 201.6 4.7 45.5 GB view
Batching Results
Batch Size TG tok/s Speedup
9.9 1.00×
11.1 1.12×
15.1 1.53×