← Back to community benchmarks

gemma-4-26b-a4b-it

M4 Max (40c) · 128 GB · 4bit · 2026-04-03
Performance
32k
tokens
1,204
PP tok/s
45.0
TG tok/s
27211
TTFT (ms)
17.2
Peak mem (GB)
Hardware
Chip M4 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.2
macOS macOS 15.7.5
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,437 70.0 15.3 GB view
4k 1,432 66.5 15.9 GB view
8k 1,398 61.2 16.1 GB view
16k 1,334 54.4 16.5 GB view
32k 1,204 45.0 17.2 GB current
64k 992.4 28.2 18.9 GB view
128k 702.4 1.4 22.5 GB view
Batching Results
Batch Size TG tok/s Speedup
70.0 1.00×
103.3 1.48×
155.1 2.22×