← Back to community benchmarks

SuperGemma4-31b-abliterated

M5 Max (40c) · 128 GB · 4bit · 2026-05-01

Performance

64k

tokens

250.8

PP tok/s

8.3

TG tok/s

261257

TTFT (ms)

45.0

Peak mem (GB)

Hardware

Chip M5 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.3.8

macOS macOS 26.4.1

Context 65,536

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	662.1	18.1	31.3 GB	view
4k	559.1	23.6	34.3 GB	view
8k	535.5	21.5	34.8 GB	view
16k	484.2	20.8	36.0 GB	view
32k	438.7	15.0	39.0 GB	view
64k	250.8	8.3	45.0 GB	current
128k	167.3	5.5	56.9 GB	view