← Back to community benchmarks

gemma-4-26b-a4b-it

M4 (10c) · 24 GB · 4bit · 2026-04-29
Performance
8k
tokens
328.0
PP tok/s
26.8
TG tok/s
24978
TTFT (ms)
14.8
Peak mem (GB)
Hardware
Chip M4 (10c)
Memory 24 GB
GPU Cores 10
Software
oMLX v0.3.8rc1
macOS macOS 26.4.1
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 329.9 34.3 14.2 GB view
4k 333.9 31.9 14.7 GB view
8k 328.0 26.8 14.8 GB current
16k 308.7 20.8 15.2 GB view
32k 272.1 8.7 15.9 GB view
64k 183.0 4.3 17.7 GB view
Batching Results
Batch Size TG tok/s Speedup
34.3 1.00×
47.3 1.38×