← Back to community benchmarks

gemma-4-26b-a4b

M3 Ultra (60c) · 256 GB · 8bit · 2026-05-06
Performance
128k
tokens
886.2
PP tok/s
30.0
TG tok/s
147899
TTFT (ms)
32.7
Peak mem (GB)
Hardware
Chip M3 Ultra (60c)
Memory 256 GB
GPU Cores 60
Software
oMLX v0.3.8
macOS macOS 26.4.1
Context 131,072
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,291 83.4 25.8 GB view
4k 1,655 80.6 26.4 GB view
8k 1,690 76.8 26.6 GB view
16k 1,642 71.3 27.1 GB view
32k 1,496 60.3 28.0 GB view
64k 1,232 45.8 30.1 GB view
128k 886.2 30.0 32.7 GB current
Batching Results
Batch Size TG tok/s Speedup
83.4 1.00×
120.1 1.44×
165.8 1.99×