← Back to community benchmarks

gemma-4-26b-a4b-it-UD

M4 Pro (20c) · 64 GB · 4bit · 2026-04-17
Performance
64k
tokens
482.6
PP tok/s
36.3
TG tok/s
135802
TTFT (ms)
20.5
Peak mem (GB)
Hardware
Chip M4 Pro (20c)
Memory 64 GB
GPU Cores 20
Software
oMLX v0.3.6
macOS macOS 26.3
Context 65,536
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 648.5 56.3 14.7 GB view
4k 728.5 55.1 15.5 GB view
8k 681.7 52.2 15.9 GB view
16k 592.0 49.1 16.6 GB view
32k 545.8 43.8 17.9 GB view
64k 482.6 36.3 20.5 GB current
Batching Results
Batch Size TG tok/s Speedup
56.3 1.00×
86.2 1.53×
103.4 1.84×
120.5 2.14×