← Back to community benchmarks

gemma-4-26b-a4b-it-UD

M5 Max (40c) · 128 GB · 4bit · 2026-05-20
Performance
8k
tokens
3,566
PP tok/s
95.6
TG tok/s
2298
TTFT (ms)
15.5
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.9rc1
macOS macOS 26.5
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 2,301 103.6 14.7 GB view
4k 3,342 100.3 15.4 GB view
8k 3,566 95.6 15.5 GB current
16k 3,126 87.8 16.0 GB view
32k 2,688 75.1 16.9 GB view
64k 2,161 58.4 19.0 GB view
128k 1,471 38.7 23.1 GB view
195k 869.0 25.2 28.1 GB view
Batching Results
Batch Size TG tok/s Speedup
103.6 1.00×
157.3 1.52×
215.3 2.08×
189.3 1.83×