← Back to community benchmarks

gemma-4-26b-a4b-it

M5 Max (40c) · 128 GB · 4bit · 2026-05-02
Performance
195k
tokens
1,025
PP tok/s
25.1
TG tok/s
195091
TTFT (ms)
25.0
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.8
macOS macOS 26.4.1
Context 200,000
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,431 117.3 14.3 GB view
4k 2,593 113.2 14.9 GB view
8k 3,722 107.0 15.1 GB view
16k 3,272 96.3 15.6 GB view
32k 2,849 79.4 16.0 GB view
64k 2,269 59.1 17.7 GB view
128k 1,519 38.3 21.2 GB view
195k 1,025 25.1 25.0 GB current
Batching Results
Batch Size TG tok/s Speedup
117.3 1.00×
130.9 1.12×
176.7 1.51×