← Back to community benchmarks

gemma-4-26b-a4b-it

M5 Pro (20c) · 64 GB · 6bit · 2026-06-05
Performance
32k
tokens
1,595
PP tok/s
39.4
TG tok/s
20546
TTFT (ms)
21.7
Peak mem (GB)
Hardware
Chip M5 Pro (20c)
Memory 64 GB
GPU Cores 20
Software
oMLX v0.3.8
macOS macOS 26.5.1
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,345 55.8 20.0 GB view
4k 1,786 53.9 20.7 GB view
8k 1,870 51.2 20.8 GB view
16k 1,802 46.6 21.3 GB view
32k 1,595 39.4 21.7 GB current
64k 1,264 30.5 23.4 GB view
128k 786.8 19.4 27.0 GB view
195k 554.0 15.1 30.7 GB view
Batching Results
Batch Size TG tok/s Speedup
55.8 1.00×
71.5 1.28×
87.1 1.56×
104.7 1.88×