← Back to community benchmarks

nightmedia-Qwen3-Coder-Next-mxfp8

M4 Max (40c) · 128 GB · 8bit · 2026-05-05
Performance
4k
tokens
1,257
PP tok/s
64.6
TG tok/s
3258
TTFT (ms)
78.7
Peak mem (GB)
Hardware
Chip M4 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.8
macOS macOS 26.4.1
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,054 67.2 77.8 GB view
4k 1,257 64.6 78.7 GB current
Batching Results
Batch Size TG tok/s Speedup
67.2 1.00×
99.3 1.48×
129.9 1.93×