← Back to community benchmarks

gemma-4-31B-it-oQ4

M5 Max (40c) · 128 GB · 4bit · 2026-05-04
Performance
16k
tokens
499.8
PP tok/s
19.2
TG tok/s
32782
TTFT (ms)
22.4
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.8
macOS macOS 26.4
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 655.9 25.2 18.1 GB view
4k 549.9 22.8 20.7 GB view
8k 518.8 21.4 21.3 GB view
16k 499.8 19.2 22.4 GB current
32k 443.6 16.1 25.3 GB view
64k 347.5 11.7 31.3 GB view
128k 256.9 8.7 43.4 GB view
195k 200.6 6.4 58.4 GB view