← Back to community benchmarks

Mirua_v1

M3 Ultra (60c) · 256 GB · 4bit · 2026-05-02
Performance
16k
tokens
2,572
PP tok/s
51.8
TG tok/s
6371
TTFT (ms)
22.5
Peak mem (GB)
Hardware
Chip M3 Ultra (60c)
Memory 256 GB
GPU Cores 60
Software
oMLX v0.3.8
macOS macOS 26.4.1
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,563 99.6 20.3 GB view
4k 2,214 97.5 21.0 GB view
8k 2,265 92.6 21.4 GB view
16k 2,572 51.8 22.5 GB current
32k 2,845 50.6 23.2 GB view
64k 2,484 48.5 24.1 GB view
128k 1,890 44.9 26.1 GB view
195k 1,446 42.1 29.1 GB view
Batching Results
Batch Size TG tok/s Speedup
99.6 1.00×
171.2 1.72×
262.5 2.64×
373.2 3.75×