← Back to community benchmarks

Qwen3-Coder-Next

M4 Max (40c) · 128 GB · 6bit · 2026-04-26
Performance
32k
tokens
1,057
PP tok/s
52.7
TG tok/s
30994
TTFT (ms)
64.9
Peak mem (GB)
Hardware
Chip M4 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.6
macOS macOS 26.4.1
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,110 76.6 61.6 GB view
4k 1,304 73.8 62.5 GB view
8k 1,290 70.4 62.8 GB view
16k 1,211 62.5 63.5 GB view
32k 1,057 52.7 64.9 GB current
64k 830.5 36.4 67.7 GB view
128k 559.6 18.9 73.3 GB view
Batching Results
Batch Size TG tok/s Speedup
76.6 1.00×
111.5 1.46×
150.5 1.96×
192.9 2.52×