← Back to community benchmarks

gemma-4-26B-A4B-it5bit

M1 Max (32c) · 32 GB · 5bit · 2026-04-25
Performance
8k
tokens
480.4
PP tok/s
34.9
TG tok/s
17053
TTFT (ms)
17.7
Peak mem (GB)
Hardware
Chip M1 Max (32c)
Memory 32 GB
GPU Cores 32
Software
oMLX v0.3.7
macOS macOS 26.4
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 389.5 42.0 17.1 GB view
4k 455.5 39.6 17.6 GB view
8k 480.4 34.9 17.7 GB current
16k 470.4 24.8 18.1 GB view
32k 443.1 12.7 18.9 GB view
64k 377.2 8.0 20.6 GB view
Batching Results
Batch Size TG tok/s Speedup
42.0 1.00×
57.4 1.37×
69.9 1.66×