← Back to community benchmarks

Laguna-S2.1-oQ6e

M3 Ultra (80c) · 512 GB · 6bit · 2026-07-22

Performance

16k

tokens

791.3

PP tok/s

43.0

TG tok/s

20707

TTFT (ms)

88.2

Peak mem (GB)

Hardware

Chip M3 Ultra (80c)

Memory 512 GB

GPU Cores 80

Software

oMLX v0.5.3

macOS macOS 26.5.2

Context 16,384

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	848.8	51.5	87.0 GB	view
4k	858.5	50.4	87.3 GB	view
8k	829.5	49.4	87.6 GB	view
16k	791.3	43.0	88.2 GB	current
32k	717.0	42.1	89.3 GB	view
64k	577.9	37.8	91.7 GB	view
128k	442.2	30.5	96.4 GB	view