← Back to community benchmarks

MiniMax-M2

M3 Ultra (80c) · 512 GB · 8bit · 2026-03-10
Performance
16k
tokens
560.8
PP tok/s
21.3
TG tok/s
29216
TTFT (ms)
231.1
Peak mem (GB)
Hardware
Chip M3 Ultra (80c)
Memory 512 GB
GPU Cores 80
Software
oMLX v0.2.7
macOS macOS 15.7.4
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 583.3 34.2 227.2 GB view
4k 705.0 30.2 228.2 GB view
8k 660.0 26.3 229.1 GB view
16k 560.8 21.3 231.1 GB current
32k 423.2 14.9 235.0 GB view
64k 282.3 9.6 242.8 GB view
Batching Results
Batch Size TG tok/s Speedup
34.2 1.00×
50.0 1.46×
65.5 1.92×
98.4 2.88×