← Back to community benchmarks
Qwen3-Coder-30B-A3B-Instruct
Performance
4k
tokens
368.3
PP tok/s
35.8
TG tok/s
11122
TTFT (ms)
17.0
Peak mem (GB)
Hardware
Chip
M4 (10c)
Memory
32 GB
GPU Cores
10
Software
oMLX
v0.2.7
macOS
macOS 26.2
Context
4,096