← Back to community benchmarks

Qwen3-Coder-Next-mxfp4

M2 Ultra (60c) · 128 GB · 4bit · 2026-03-10
Performance
4k
tokens
621.5
PP tok/s
37.1
TG tok/s
6590
TTFT (ms)
40.9
Peak mem (GB)
Hardware
Chip M2 Ultra (60c)
Memory 128 GB
GPU Cores 60
Software
oMLX v0.2.7
macOS macOS 26.3
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 530.9 38.1 40.7 GB view
4k 621.5 37.1 40.9 GB current
Batching Results
Batch Size TG tok/s Speedup
1× 38.1 1.00×
2× 63.2 1.66×
4× 100.5 2.64×