← Back to community benchmarks

gemma-4-26B-A4B-it-qat

M3 Pro (14c) · 36 GB · 4bit · 2026-07-25
Performance
32k
tokens
372.5
PP tok/s
29.8
TG tok/s
87978
TTFT (ms)
16.0
Peak mem (GB)
Hardware
Chip M3 Pro (14c)
Memory 36 GB
GPU Cores 14
Software
oMLX v0.5.3
macOS macOS 27.0
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 473.3 38.7 14.3 GB view
4k 465.0 35.1 14.7 GB view
8k 448.1 35.9 14.8 GB view
16k 421.8 33.8 15.2 GB view
32k 372.5 29.8 16.0 GB current
64k 304.9 24.6 17.7 GB view
Batching Results
Batch Size TG tok/s Speedup
38.7 1.00×
79.4 2.05×
166.3 4.30×