← Back to community benchmarks

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

M5 Pro (20c) · 64 GB · 4bit · 2026-05-12
Performance
32k
tokens
1,845
PP tok/s
46.9
TG tok/s
17762
TTFT (ms)
17.9
Peak mem (GB)
Hardware
Chip M5 Pro (20c)
Memory 64 GB
GPU Cores 20
Software
oMLX v0.3.9.dev1
macOS macOS 26.4.1
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,518 67.6 14.7 GB view
4k 2,075 65.1 15.6 GB view
8k 2,145 61.7 15.9 GB view
16k 2,085 55.0 16.6 GB view
32k 1,845 46.9 17.9 GB current
Batching Results
Batch Size TG tok/s Speedup
67.6 1.00×
100.2 1.48×
127.5 1.89×