← Back to community benchmarks

gemma-4-26b-a4b-it

M5 Max (32c) · 36 GB · 4bit · 2026-05-27
Performance
195k
tokens
479.0
PP tok/s
20.5
TG tok/s
417543
TTFT (ms)
27.6
Peak mem (GB)
Hardware
Chip M5 Max (32c)
Memory 36 GB
GPU Cores 32
Software
oMLX v0.3.10
macOS macOS 26.5
Context 200,000
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,840 96.8 14.3 GB view
4k 2,700 92.7 15.0 GB view
8k 2,885 88.4 15.1 GB view
16k 2,742 80.2 15.6 GB view
32k 2,275 62.4 16.5 GB view
64k 1,616 50.7 18.5 GB view
128k 841.3 25.5 22.7 GB view
195k 479.0 20.5 27.6 GB current
Batching Results
Batch Size TG tok/s Speedup
96.8 1.00×
100.5 1.04×
121.5 1.26×