← Back to community benchmarks

gemma-4-26b-a4b-it

M5 Max (40c) · 128 GB · 4bit · 2026-05-15
Performance
128k
tokens
1,274
PP tok/s
36.9
TG tok/s
102890
TTFT (ms)
22.7
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.9.dev1
macOS macOS 26.4.1
Context 131,072
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,886 108.1 14.3 GB view
4k 2,963 99.3 14.9 GB view
8k 3,333 98.5 15.1 GB view
16k 2,955 89.1 15.6 GB view
32k 2,522 75.8 16.5 GB view
64k 1,945 54.4 18.5 GB view
128k 1,274 36.9 22.7 GB current
Batching Results
Batch Size TG tok/s Speedup
108.1 1.00×
132.3 1.22×
153.6 1.42×