← Back to community benchmarks
Qwen3-Coder-30B-A3B-Instruct-oQ3
Performance
16k
tokens
907.0
PP tok/s
27.4
TG tok/s
18064
TTFT (ms)
15.1
Peak mem (GB)
Hardware
Chip
M5 (10c)
Memory
32 GB
GPU Cores
10
Software
oMLX
v0.3.7
macOS
macOS 26.4.1
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,080 | 56.8 | 13.6 GB | view |
| 4k | 1,310 | 47.5 | 13.9 GB | view |
| 8k | 1,152 | 38.1 | 14.3 GB | view |
| 16k | 907.0 | 27.4 | 15.1 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 56.8 | 1.00× |
| 2× | 74.9 | 1.32× |
| 4× | 90.3 | 1.59× |