← Back to community benchmarks

gemma-4-26B-A4B-it

M3 Ultra (60c) · 256 GB · 8bit · 2026-05-20
Performance
32k
tokens
1,484
PP tok/s
61.1
TG tok/s
22082
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 1,334 85.1 25.8 GB view
4k 1,666 81.5 26.4 GB view
8k 1,701 77.6 26.6 GB view
16k 1,648 72.0 27.1 GB view
32k 1,484 61.1 28.0 GB current
64k 1,213 46.3 30.0 GB view
Batching Results
Batch Size TG tok/s Speedup
85.1 1.00×
125.5 1.47×
169.0 1.99×
234.4 2.75×