← Back to community benchmarks

gemma-4-26b-a4b-it

M5 Pro (20c) · 64 GB · 4bit · 2026-05-13
Performance
16k
tokens
2,007
PP tok/s
59.6
TG tok/s
8162
TTFT (ms)
15.6
Peak mem (GB)
Hardware
Chip M5 Pro (20c)
Memory 64 GB
GPU Cores 20
Software
oMLX v0.3.8
macOS macOS 26.4.1
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,511 74.5 14.3 GB view
4k 1,955 71.3 14.9 GB view
8k 2,061 66.8 15.1 GB view
16k 2,007 59.6 15.6 GB current
32k 1,806 48.6 16.5 GB view
64k 1,391 34.5 18.5 GB view
Batching Results
Batch Size TG tok/s Speedup
74.5 1.00×
104.2 1.40×
135.1 1.81×
155.4 2.09×