← Back to community benchmarks
Qwen3-VL-30B-A3B-Instruct
Performance
4k
tokens
356.0
PP tok/s
27.6
TG tok/s
11504
TTFT (ms)
18.1
Peak mem (GB)
Hardware
Chip
M4 (10c)
Memory
32 GB
GPU Cores
10
Software
oMLX
v0.2.6
macOS
macOS 26.3.1
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 416.6 | 34.5 | 17.5 GB | view |
| 4k | 356.0 | 27.6 | 18.1 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 34.5 | 1.00× |
| 2× | 51.3 | 1.49× |
| 4× | 51.9 | 1.50× |