← Back to community benchmarks

gemma4-moe-26b-a4b

M5 Max (40c) · 128 GB · bf16 · 2026-05-10
Performance
16k
tokens
3,425
PP tok/s
52.8
TG tok/s
4783
TTFT (ms)
49.1
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.9.dev1
macOS macOS 26.4.1
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 2,014 58.2 47.8 GB view
4k 3,041 56.6 48.5 GB view
8k 3,396 55.2 48.6 GB view
16k 3,425 52.8 49.1 GB current
Batching Results
Batch Size TG tok/s Speedup
58.2 1.00×
59.6 1.02×
101.9 1.75×
119.9 2.06×