← Back to community benchmarks

gemma-4-E4B-it-heretic-oQ6

M3 (10c) · 16 GB · 6bit · 2026-05-19
Performance
32k
tokens
358.0
PP tok/s
11.9
TG tok/s
91527
TTFT (ms)
8.0
Peak mem (GB)
Hardware
Chip M3 (10c)
Memory 16 GB
GPU Cores 10
Software
oMLX v0.3.9.dev2
macOS macOS 26.5
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 525.0 32.7 6.9 GB view
4k 527.7 30.6 7.0 GB view
8k 501.3 28.0 7.2 GB view
16k 456.4 24.0 7.4 GB view
32k 358.0 11.9 8.0 GB current
64k 258.2 9.1 9.2 GB view
Batching Results
Batch Size TG tok/s Speedup
32.7 1.00×
15.6 0.48×
23.7 0.72×
26.8 0.82×