← Back to community benchmarks

gemma-4-26b-a4b-it

M4 Max (40c) · 128 GB · 8bit · 2026-05-14
Performance
64k
tokens
1,024
PP tok/s
50.9
TG tok/s
63983
TTFT (ms)
30.0
Peak mem (GB)
Hardware
Chip M4 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.8
macOS macOS 26.4.1
Context 65,536
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,195 77.5 25.8 GB view
4k 1,405 75.2 26.3 GB view
8k 1,415 73.2 26.7 GB view
16k 1,362 68.7 27.1 GB view
32k 1,234 61.2 28.0 GB view
64k 1,024 50.9 30.0 GB current
Batching Results
Batch Size TG tok/s Speedup
77.5 1.00×
115.7 1.49×
158.2 2.04×
201.8 2.60×