← Back to community benchmarks

gemma-4-26b-a4b-it-UD

M5 Max (40c) · 128 GB · 4bit · 2026-05-11
Performance
16k
tokens
3,441
PP tok/s
88.5
TG tok/s
4761
TTFT (ms)
16.7
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.9.dev1
macOS macOS 26.5
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 2,197 104.7 15.2 GB view
4k 3,218 99.5 15.9 GB view
8k 3,559 94.5 16.2 GB view
16k 3,441 88.5 16.7 GB current
32k 3,133 76.4 17.6 GB view
64k 2,425 59.5 21.8 GB view
128k 1,625 42.0 25.1 GB view
Batching Results
Batch Size TG tok/s Speedup
104.7 1.00×
158.6 1.51×
216.6 2.07×
420.6 4.02×