← Back to community benchmarks

GLM-4.6V-5bit

M3 Ultra (80c) · 512 GB · 5bit · 2026-03-10
Performance
4k
tokens
654.1
PP tok/s
33.8
TG tok/s
6262
TTFT (ms)
71.8
Peak mem (GB)
Hardware
Chip M3 Ultra (80c)
Memory 512 GB
GPU Cores 80
Software
oMLX v0.2.6
macOS macOS 26.3
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 618.2 37.5 71.1 GB view
4k 654.1 33.8 71.8 GB current
8k 591.7 29.7 72.6 GB view
16k 481.8 24.3 74.0 GB view
32k 350.2 17.8 76.9 GB view
64k 226.2 11.6 82.8 GB view
Batching Results
Batch Size TG tok/s Speedup
37.5 1.00×
53.2 1.42×
74.1 1.98×
83.7 2.23×