← Back to community benchmarks

gemma-4-26b-a4b-it

M5 Max (40c) · 128 GB · 4bit · 2026-05-17
Performance
195k
tokens
1,186
PP tok/s
28.4
TG tok/s
168644
TTFT (ms)
27.5
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 200,000
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 2,045 110.4 14.3 GB view
4k 3,374 106.4 14.9 GB view
8k 3,678 103.5 15.1 GB view
16k 3,634 93.4 15.6 GB view
32k 3,166 78.8 16.5 GB view
64k 2,383 59.2 18.5 GB view
128k 1,652 41.7 22.6 GB view
195k 1,186 28.4 27.5 GB current
Batching Results
Batch Size TG tok/s Speedup
110.4 1.00×
147.3 1.33×
188.5 1.71×
206.1 1.87×