← Back to community benchmarks

GLM-5

M3 Ultra (80c) · 512 GB · 4bit · 2026-03-13
Performance
16k
tokens
127.2
PP tok/s
14.1
TG tok/s
128846
TTFT (ms)
404.0
Peak mem (GB)
Hardware
Chip M3 Ultra (80c)
Memory 512 GB
GPU Cores 80
Software
oMLX v0.2.11a1
macOS macOS 15.5
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
8k 163.0 14.9 398.0 GB view
16k 127.2 14.1 404.0 GB current