← Back to community benchmarks

Gemma-4-26B-A4B-it-RAM-20GB

M5 (10c) · 32 GB · 4bit · 2026-04-16
Performance
32k
tokens
822.1
PP tok/s
8.5
TG tok/s
39858
TTFT (ms)
20.6
Peak mem (GB)
Hardware
Chip M5 (10c)
Memory 32 GB
GPU Cores 10
Software
oMLX v0.3.6
macOS macOS 26.4.1
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 753.6 22.1 18.8 GB view
4k 862.5 21.8 19.3 GB view
8k 978.9 21.0 19.4 GB view
32k 822.1 8.5 20.6 GB current
Batching Results
Batch Size TG tok/s Speedup
22.1 1.00×
28.9 1.31×