← Back to community benchmarks

gemma-4-26B-A4B-it

M3 Pro (18c) · 36 GB · 4bit · 2026-05-09
Performance
32k
tokens
483.0
PP tok/s
31.6
TG tok/s
67848
TTFT (ms)
16.0
Peak mem (GB)
Hardware
Chip M3 Pro (18c)
Memory 36 GB
GPU Cores 18
Software
oMLX v0.3.8
macOS macOS 26.4.1
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 523.8 41.8 14.3 GB view
4k 563.1 40.6 14.9 GB view
8k 568.0 39.5 15.1 GB view
16k 539.0 36.4 15.6 GB view
32k 483.0 31.6 16.0 GB current
64k 395.2 25.2 17.7 GB view
Batching Results
Batch Size TG tok/s Speedup
41.8 1.00×
62.4 1.49×
78.6 1.88×