← Back to community benchmarks

gemma-4-26b-a4b-it

M5 Pro (20c) · 64 GB · 4bit · 2026-05-19
Performance
195k
tokens
725.5
PP tok/s
15.4
TG tok/s
275672
TTFT (ms)
25.0
Peak mem (GB)
Hardware
Chip M5 Pro (20c)
Memory 64 GB
GPU Cores 20
Software
oMLX v0.3.8
macOS macOS 26.5
Context 200,000
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,550 75.0 14.3 GB view
4k 2,017 72.5 14.9 GB view
8k 2,103 67.8 15.1 GB view
16k 2,047 59.5 15.6 GB view
32k 1,829 47.8 16.0 GB view
64k 1,407 35.4 17.7 GB view
128k 978.9 23.8 21.2 GB view
195k 725.5 15.4 25.0 GB current
Batching Results
Batch Size TG tok/s Speedup
75.0 1.00×
107.4 1.43×
139.8 1.86×
166.5 2.22×