← Back to community benchmarks

gemma-4-26b-a4b-it

M5 Pro (20c) · 64 GB · 8bit · 2026-05-27
Performance
16k
tokens
1,853
PP tok/s
43.1
TG tok/s
8843
TTFT (ms)
26.9
Peak mem (GB)
Hardware
Chip M5 Pro (20c)
Memory 64 GB
GPU Cores 20
Software
oMLX v0.3.12
macOS macOS 26.5
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,371 50.5 25.8 GB view
16k 1,853 43.1 26.9 GB current
64k 1,318 29.0 29.7 GB view
128k 857.2 20.6 32.0 GB view
Batching Results
Batch Size TG tok/s Speedup
50.5 1.00×
72.0 1.43×
94.2 1.87×