← Back to community benchmarks

ornith-1.0-35b-abliterated-mxfp4-vision-mtplx

M3 Ultra (60c) · 256 GB · 4bit · 2026-07-23
Performance
16k
tokens
1,956
PP tok/s
91.5
TG tok/s
8377
TTFT (ms)
19.3
Peak mem (GB)
Hardware
Chip M3 Ultra (60c)
Memory 256 GB
GPU Cores 60
Software
oMLX v0.5.3
macOS macOS 26.5.2
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 2,176 106.5 18.4 GB view
4k 2,312 104.5 19.1 GB view
8k 2,182 97.1 19.2 GB view
16k 1,956 91.5 19.3 GB current
32k 1,627 85.8 19.7 GB view
64k 1,225 73.3 20.5 GB view
Batching Results
Batch Size TG tok/s Speedup
106.5 1.00×
197.2 1.85×
286.1 2.69×
405.3 3.81×