← Back to community benchmarks

gemma4-26b

M4 Pro (20c) · 48 GB · 4bit · 2026-05-15
Performance
195k
tokens
294.4
PP tok/s
3.3
TG tok/s
679309
TTFT (ms)
34.5
Peak mem (GB)
Hardware
Chip M4 Pro (20c)
Memory 48 GB
GPU Cores 20
Software
oMLX v0.3.8
macOS macOS 26.4.1
Context 200,000
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 684.7 67.6 14.3 GB view
4k 751.9 65.5 14.8 GB view
8k 746.1 62.6 15.2 GB view
16k 693.2 57.3 15.6 GB view
32k 598.4 50.7 16.5 GB view
64k 510.5 40.5 18.5 GB view
128k 382.7 28.7 22.7 GB view
195k 294.4 3.3 34.5 GB current
Batching Results
Batch Size TG tok/s Speedup
67.6 1.00×
33.5 0.50×
72.3 1.07×
288.9 4.27×