← Back to community benchmarks

Huihui-Qwen3-Next-80B-A3B-Instruct-abliterated

M4 Max (40c) · 128 GB · 4bit · 2026-03-10
Performance
16k
tokens
1,214
PP tok/s
64.9
TG tok/s
13491
TTFT (ms)
43.9
Peak mem (GB)
Hardware
Chip M4 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.2.6
macOS macOS 26.4
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,147 88.8 43.0 GB view
4k 1,309 84.2 43.2 GB view
8k 1,296 80.0 43.4 GB view
16k 1,214 64.9 43.9 GB current
32k 1,053 40.1 44.7 GB view
64k 830.1 26.7 46.5 GB view
Batching Results
Batch Size TG tok/s Speedup
88.8 1.00×
147.8 1.66×
209.8 2.36×
281.6 3.17×