← Back to community benchmarks

gemma-4-26b-a4b-it

M3 Max (30c) · 36 GB · 4bit · 2026-05-12
Performance
32k
tokens
701.9
PP tok/s
48.2
TG tok/s
46687
TTFT (ms)
16.5
Peak mem (GB)
Hardware
Chip M3 Max (30c)
Memory 36 GB
GPU Cores 30
Software
oMLX v0.3.9.dev2
macOS macOS 26.5
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 833.2 72.1 14.3 GB view
4k 934.9 69.4 14.9 GB view
8k 934.5 65.7 15.1 GB view
16k 814.6 59.8 15.6 GB view
32k 701.9 48.2 16.5 GB current
64k 522.5 20.2 18.5 GB view
Batching Results
Batch Size TG tok/s Speedup
72.1 1.00×
43.6 0.60×
48.4 0.67×