← Back to community benchmarks

gemma-4-26b-a4b-it

M4 Max (40c) · 64 GB · 8bit · 2026-05-13
Performance
16k
tokens
1,299
PP tok/s
68.7
TG tok/s
12611
TTFT (ms)
27.1
Peak mem (GB)
Hardware
Chip M4 Max (40c)
Memory 64 GB
GPU Cores 40
Software
oMLX v0.3.9.dev2
macOS macOS 26.5
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,164 78.2 25.8 GB view
4k 1,395 75.1 26.3 GB view
8k 1,361 72.6 26.8 GB view
16k 1,299 68.7 27.1 GB current
32k 1,092 60.3 28.0 GB view
64k 939.7 49.9 30.0 GB view
128k 667.8 36.9 34.2 GB view
Batching Results
Batch Size TG tok/s Speedup
78.2 1.00×
118.1 1.51×
158.2 2.02×