← Back to community benchmarks

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

M4 Max (40c) · 128 GB · 8bit · 2026-06-05
Performance
16k
tokens
1,313
PP tok/s
55.0
TG tok/s
12481
TTFT (ms)
28.8
Peak mem (GB)
Hardware
Chip M4 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.4.2.dev2
macOS macOS 15.7.4
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,130 71.1 25.8 GB view
4k 1,428 67.0 26.3 GB view
8k 1,377 62.9 27.1 GB view
16k 1,313 55.0 28.8 GB current
32k 1,105 41.9 32.3 GB view
64k 752.4 27.4 39.3 GB view