← Back to community benchmarks

gemma-4-26b-a4b-it

M5 Pro (20c) · 64 GB · 4bit · 2026-04-17
Performance
8k
tokens
2,065
PP tok/s
58.7
TG tok/s
3967
TTFT (ms)
14.8
Peak mem (GB)
Hardware
Chip M5 Pro (20c)
Memory 64 GB
GPU Cores 20
Software
oMLX v0.3.6
macOS macOS 26.4.1
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,535 74.3 14.2 GB view
4k 1,982 70.5 14.7 GB view
8k 2,065 58.7 14.8 GB current
16k 2,025 44.2 15.2 GB view
Batching Results
Batch Size TG tok/s Speedup
74.3 1.00×
105.8 1.42×
137.0 1.84×