← Back to community benchmarks
Qwen3-30B-A3B-Instruct-2507
Performance
16k
tokens
957.0
PP tok/s
55.8
TG tok/s
17121
TTFT (ms)
18.2
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
48 GB
GPU Cores
40
Software
oMLX
v0.2.7
macOS
macOS 26.3.1
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,280 | 113.3 | 16.6 GB | view |
| 4k | 1,398 | 92.6 | 17.1 GB | view |
| 8k | 1,238 | 75.8 | 17.4 GB | view |
| 16k | 957.0 | 55.8 | 18.2 GB | current |
| 32k | 646.2 | 37.2 | 19.7 GB | view |
| 64k | 390.9 | 22.8 | 22.7 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 113.3 | 1.00× |
| 2× | 165.7 | 1.46× |