← Back to community benchmarks

gemma-4-26b-a4b-it

M1 Max (24c) · 64 GB · bf16 · 2026-04-05
Performance
16k
tokens
615.1
PP tok/s
22.2
TG tok/s
26637
TTFT (ms)
48.7
Peak mem (GB)
Hardware
Chip M1 Max (24c)
Memory 64 GB
GPU Cores 24
Software
oMLX v0.3.4
macOS macOS 26.4
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 573.2 31.4 47.7 GB view
4k 642.1 29.9 48.2 GB view
8k 643.8 27.1 48.3 GB view
16k 615.1 22.2 48.7 GB current
32k 551.5 9.7 49.5 GB view
Batching Results
Batch Size TG tok/s Speedup
31.4 1.00×
24.8 0.79×
41.3 1.32×