← Back to community benchmarks

gemma-4-26B-A4B-it-oQ8

M5 Max (40c) · 128 GB · 8bit · 2026-06-03
Performance
64k
tokens
2,307
PP tok/s
49.9
TG tok/s
28406
TTFT (ms)
32.1
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.4.1
macOS macOS 26.6
Context 65,536
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 2,100 79.1 27.9 GB view
4k 2,961 77.3 28.5 GB view
8k 3,262 74.7 28.7 GB view
16k 3,211 69.0 29.2 GB view
32k 2,899 63.4 30.1 GB view
64k 2,307 49.9 32.1 GB current
Batching Results
Batch Size TG tok/s Speedup
79.1 1.00×
117.4 1.48×
163.8 2.07×