← Back to community benchmarks

Huihui-Qwen3.5-35B-A3B-abliterated

M3 Ultra (80c) · 512 GB · 4bit · 2026-03-09
Performance
4k
tokens
2,046
PP tok/s
82.3
TG tok/s
2002
TTFT (ms)
19.6
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 1,165 85.1 19.2 GB view
4k 2,046 82.3 19.6 GB current
Batching Results
Batch Size TG tok/s Speedup
85.1 1.00×
141.3 1.66×
244.8 2.88×