← Back to community benchmarks

gemma-4-e4b-it

M1 Ultra (48c) · 64 GB · 4bit · 2026-04-03
Performance
16k
tokens
1,199
PP tok/s
32.9
TG tok/s
13667
TTFT (ms)
5.8
Peak mem (GB)
Hardware
Chip M1 Ultra (48c)
Memory 64 GB
GPU Cores 48
Software
oMLX v0.3.2
macOS macOS 26.3.1
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,199 46.2 5.5 GB view
4k 1,227 44.8 5.6 GB view
8k 1,228 41.2 5.7 GB view
16k 1,199 32.9 5.8 GB current
32k 1,135 24.8 6.4 GB view
64k 1,026 16.7 7.5 GB view
Batching Results
Batch Size TG tok/s Speedup
46.2 1.00×
59.1 1.28×
75.7 1.64×
119.7 2.59×