← Back to community benchmarks

MiniMax-M2___5

M3 Ultra (80c) · 512 GB · 8bit · 2026-03-11
Performance
16k
tokens
574.9
PP tok/s
23.2
TG tok/s
28500
TTFT (ms)
231.1
Peak mem (GB)
Hardware
Chip M3 Ultra (80c)
Memory 512 GB
GPU Cores 80
Software
oMLX v0.2.6
macOS macOS 26.3.1
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 621.7 35.9 227.2 GB view
4k 724.9 31.8 228.2 GB view
8k 675.2 28.2 229.1 GB view
16k 574.9 23.2 231.1 GB current
Batching Results
Batch Size TG tok/s Speedup
35.9 1.00×
51.2 1.43×
64.6 1.80×