← Back to community benchmarks

unsloth-gemma-4-E4B-it-qat-oQ4

M2 Pro (16c) · 16 GB · 4bit · 2026-07-21
Performance
32k
tokens
447.7
PP tok/s
31.0
TG tok/s
73192
TTFT (ms)
5.7
Peak mem (GB)
Hardware
Chip M2 Pro (16c)
Memory 16 GB
GPU Cores 16
Software
oMLX v0.5.2rc2
macOS macOS 26.5.2
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 515.8 45.2 4.8 GB view
4k 495.7 43.0 4.9 GB view
8k 484.5 40.8 5.0 GB view
16k 471.3 36.5 5.1 GB view
32k 447.7 31.0 5.7 GB current
64k 405.1 22.8 6.6 GB view
128k 319.8 15.7 8.5 GB view
Batching Results
Batch Size TG tok/s Speedup
45.2 1.00×
57.0 1.26×