← Back to community benchmarks

gemma-4-26b-a4b-it

M5 (10c) · 32 GB · 4bit · 2026-05-19
Performance
32k
tokens
862.8
PP tok/s
27.1
TG tok/s
37980
TTFT (ms)
16.2
Peak mem (GB)
Hardware
Chip M5 (10c)
Memory 32 GB
GPU Cores 10
Software
oMLX v0.3.9.dev2
macOS macOS 26.4.1
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 879.8 40.1 14.2 GB view
4k 1,021 38.2 14.9 GB view
32k 862.8 27.1 16.2 GB current
64k 632.1 19.1 18.5 GB view
128k 450.7 12.3 22.7 GB view
Batching Results
Batch Size TG tok/s Speedup
40.1 1.00×
57.2 1.43×
71.1 1.77×