← Back to community benchmarks
GLM-4___7-Flash
Performance
32k
tokens
612.7
PP tok/s
31.7
TG tok/s
53479
TTFT (ms)
39.5
Peak mem (GB)
Hardware
Chip
M3 Ultra (80c)
Memory
512 GB
GPU Cores
80
Software
oMLX
v0.2.12
macOS
macOS 15.3
Context
32,768