← Back to community benchmarks

Huihui-Qwen3.5-4B-abliterated

M3 Pro (14c) · 36 GB · unknown · 2026-03-12
Performance
4k
tokens
545.3
PP tok/s
15.3
TG tok/s
7512
TTFT (ms)
9.6
Peak mem (GB)
Hardware
Chip M3 Pro (14c)
Memory 36 GB
GPU Cores 14
Software
oMLX v0.2.9
macOS macOS 26.3.1
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 534.5 15.5 9.4 GB view
4k 545.3 15.3 9.6 GB current
Batching Results
Batch Size TG tok/s Speedup
15.5 1.00×
17.0 1.10×
33.5 2.16×