← Back to community benchmarks

gemma-4-26B-A4B-it

M3 Ultra (60c) · 256 GB · 8bit · 2026-05-20
Performance
32k
tokens
1,482
PP tok/s
61.5
TG tok/s
22112
TTFT (ms)
28.0
Peak mem (GB)
Hardware
Chip M3 Ultra (60c)
Memory 256 GB
GPU Cores 60
Software
oMLX v0.3.9rc1
macOS macOS 26.4.1
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,316 84.5 25.8 GB view
4k 1,664 81.1 26.4 GB view
8k 1,700 78.1 26.6 GB view
16k 1,649 71.6 27.1 GB view
32k 1,482 61.5 28.0 GB current
64k 1,212 46.5 30.0 GB view
Batching Results
Batch Size TG tok/s Speedup
84.5 1.00×
129.2 1.53×
173.0 2.05×
235.6 2.79×