← Back to community benchmarks

Ornith-1.0-35B-uncensored-hereticoQ8

M5 Max (40c) · 128 GB · 8bit · 2026-07-21
Performance
8k
tokens
4,060
PP tok/s
84.5
TG tok/s
2018
TTFT (ms)
36.7
Peak mem (GB)
Hardware
Chip M5 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.5.2rc2
macOS macOS 26.5.2
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 3,518 89.2 35.5 GB view
4k 4,309 82.2 36.3 GB view
8k 4,060 84.5 36.7 GB current
16k 3,734 84.9 37.3 GB view
32k 3,103 85.0 38.6 GB view
64k 2,190 66.5 41.3 GB view
Batching Results
Batch Size TG tok/s Speedup
89.2 1.00×
151.8 1.70×
239.5 2.68×
476.4 5.34×