← Back to community benchmarks

gemma-4-26b-a4b-it

M5 Max (40c) · 128 GB · 4bit · 2026-06-02
Performance
32k
tokens
2,669
PP tok/s
64.7
TG tok/s
12280
TTFT (ms)
16.5
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.4.0rc1
macOS macOS 26.5
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 2,042 92.3 14.3 GB view
4k 3,047 91.6 14.9 GB view
8k 3,346 88.2 15.1 GB view
16k 3,051 86.4 15.8 GB view
32k 2,669 64.7 16.5 GB current
64k 2,034 52.4 18.5 GB view
Batching Results
Batch Size TG tok/s Speedup
92.3 1.00×
140.8 1.53×
119.7 1.30×
162.9 1.76×