← Back to community benchmarks

gemma-4-26b-a4b-it

M4 Max (40c) · 48 GB · 4bit · 2026-05-02
Performance
8k
tokens
1,360
PP tok/s
86.1
TG tok/s
6022
TTFT (ms)
14.8
Peak mem (GB)
Hardware
Chip M4 Max (40c)
Memory 48 GB
GPU Cores 40
Software
oMLX v0.3.7rc2
macOS macOS 26.4.1
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,109 101.8 14.3 GB view
4k 1,367 90.5 14.8 GB view
8k 1,360 86.1 14.8 GB current
16k 1,284 57.7 15.3 GB view
32k 1,106 18.6 16.0 GB view
64k 930.5 8.6 17.7 GB view
128k 623.7 4.3 21.2 GB view
195k 475.8 3.2 25.0 GB view
Batching Results
Batch Size TG tok/s Speedup
101.8 1.00×
149.5 1.47×
203.1 2.00×