← Back to community benchmarks

gemma-4-26B-A4b-it

M1 Max (32c) · 32 GB · 4bit · 2026-04-12
Performance
8k
tokens
492.1
PP tok/s
44.7
TG tok/s
16648
TTFT (ms)
14.5
Peak mem (GB)
Hardware
Chip M1 Max (32c)
Memory 32 GB
GPU Cores 32
Software
oMLX v0.3.5.dev1
macOS macOS 26.3.1
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 423.5 58.4 14.0 GB view
4k 486.1 53.8 14.4 GB view
8k 492.1 44.7 14.5 GB current
16k 481.1 33.1 14.9 GB view
32k 451.6 22.7 15.7 GB view
Batching Results
Batch Size TG tok/s Speedup
58.4 1.00×
72.9 1.25×
88.0 1.51×