← Back to community benchmarks

gemma-4-26b-a4b-it

M4 Max (40c) · 128 GB · 4bit · 2026-04-03
Performance
16k
tokens
1,301
PP tok/s
55.5
TG tok/s
12597
TTFT (ms)
15.7
Peak mem (GB)
Hardware
Chip M4 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.2
macOS macOS 15.7.5
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,400 70.5 14.6 GB view
4k 1,327 66.9 15.1 GB view
8k 1,362 61.4 15.3 GB view
16k 1,301 55.5 15.7 GB current
32k 1,179 47.5 16.5 GB view
64k 980.4 29.2 18.2 GB view
128k 694.2 0.6 22.1 GB view
Batching Results
Batch Size TG tok/s Speedup
70.5 1.00×
103.4 1.47×
154.0 2.18×