← Back to community benchmarks

gemma-3-12b-it-qat

M4 (10c) · 32 GB · 4bit · 2026-03-12
Performance
1k
tokens
139.7
PP tok/s
11.3
TG tok/s
7340
TTFT (ms)
8.3
Peak mem (GB)
Hardware
Chip M4 (10c)
Memory 32 GB
GPU Cores 10
Software
oMLX v0.2.6
macOS macOS 26.3.1
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 139.7 11.3 8.3 GB current
4k 129.2 11.8 9.0 GB view
Batching Results
Batch Size TG tok/s Speedup
11.3 1.00×
22.9 2.03×
27.8 2.46×