← Back to community benchmarks

gemma-4-26b-a4b

M5 Max (40c) · 128 GB · 6bit · 2026-04-03
Performance
8k
tokens
3,670
PP tok/s
26.2
TG tok/s
2233
TTFT (ms)
21.5
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.1
macOS macOS 26.4
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 3,365 20.6 20.9 GB view
4k 3,642 63.8 21.4 GB view
8k 3,670 26.2 21.5 GB current
16k 3,529 16.9 21.9 GB view
32k 3,105 12.6 22.7 GB view
Batching Results
Batch Size TG tok/s Speedup
20.6 1.00×
44.2 2.15×
76.4 3.71×