← Back to community benchmarks

Gemopus-4-E4B-it

M3 Ultra (80c) · 512 GB · 4bit · 2026-04-15
Performance
16k
tokens
3,379
PP tok/s
75.8
TG tok/s
4849
TTFT (ms)
4.9
Peak mem (GB)
Hardware
Chip M3 Ultra (80c)
Memory 512 GB
GPU Cores 80
Software
oMLX v0.3.5
macOS macOS 26.4.1
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,472 104.9 4.6 GB view
4k 2,707 97.0 4.7 GB view
8k 3,311 88.7 4.7 GB view
16k 3,379 75.8 4.9 GB current
Batching Results
Batch Size TG tok/s Speedup
104.9 1.00×
192.6 1.84×
295.3 2.82×