← Back to community benchmarks

Qwen3-Next-Coder4.5bit

M3 Max (40c) · 128 GB · 4bit · 2026-04-15
Performance
32k
tokens
824.2
PP tok/s
41.2
TG tok/s
39756
TTFT (ms)
50.0
Peak mem (GB)
Hardware
Chip M3 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.5.dev1
macOS macOS 26.5
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 911.5 58.3 46.7 GB view
8k 969.9 46.1 48.0 GB view
32k 824.2 41.2 50.0 GB current
Batching Results
Batch Size TG tok/s Speedup
58.3 1.00×
99.5 1.71×