← Back to community benchmarks

Huihui-Qwen3.6-35B-A3B-abliterated

M3 Ultra (60c) · 96 GB · 4bit · 2026-04-27
Performance
4k
tokens
2,045
PP tok/s
92.0
TG tok/s
2003
TTFT (ms)
20.0
Peak mem (GB)
Hardware
Chip M3 Ultra (60c)
Memory 96 GB
GPU Cores 60
Software
oMLX v0.3.6
macOS macOS 26.3.1
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,397 95.8 19.3 GB view
4k 2,045 92.0 20.0 GB current
Batching Results
Batch Size TG tok/s Speedup
95.8 1.00×
162.7 1.70×
250.3 2.61×