← Back to community benchmarks

gemma-4-31B-it

M3 Ultra (80c) · 256 GB · 8bit · 2026-05-28
Performance
16k
tokens
309.2
PP tok/s
15.1
TG tok/s
52991
TTFT (ms)
36.1
Peak mem (GB)
Hardware
Chip M3 Ultra (80c)
Memory 256 GB
GPU Cores 80
Software
oMLX v0.3.12
macOS macOS 26.5
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 310.0 16.9 31.8 GB view
4k 325.3 16.9 34.5 GB view
8k 323.3 16.3 35.0 GB view
16k 309.2 15.1 36.1 GB current
32k 282.4 13.0 39.0 GB view
64k 241.3 10.6 45.1 GB view
Batching Results
Batch Size TG tok/s Speedup
16.9 1.00×
26.5 1.57×
37.6 2.22×
46.0 2.72×