← Back to community benchmarks

Qwen3-Next-Codermixed-4.5-bit

M1 Max (32c) · 64 GB · 4bit · 2026-03-18
Performance
8k
tokens
419.8
PP tok/s
37.6
TG tok/s
19512
TTFT (ms)
48.1
Peak mem (GB)
Hardware
Chip M1 Max (32c)
Memory 64 GB
GPU Cores 32
Software
oMLX v0.2.18
macOS macOS 26.3
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 393.2 42.5 46.7 GB view
4k 441.4 40.3 47.7 GB view
8k 419.8 37.6 48.1 GB current
16k 397.6 33.8 48.8 GB view
Batching Results
Batch Size TG tok/s Speedup
42.5 1.00×
64.4 1.52×
83.4 1.96×
99.2 2.33×