← Back to community benchmarks
Qwen3.5-9B
Performance
16k
tokens
414.7
PP tok/s
21.6
TG tok/s
39505
TTFT (ms)
11.7
Peak mem (GB)
Hardware
Chip
M4 Pro (20c)
Memory
24 GB
GPU Cores
20
Software
oMLX
v0.2.6
macOS
macOS 26.3.1
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 430.4 | 27.2 | 10.7 GB | view |
| 4k | 441.1 | 26.4 | 11.0 GB | view |
| 8k | 434.4 | 24.8 | 11.2 GB | view |
| 16k | 414.7 | 21.6 | 11.7 GB | current |
| 32k | 358.1 | 18.0 | 12.7 GB | view |
| 64k | 288.1 | 10.7 | 14.8 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 27.2 | 1.00× |
| 2× | 30.8 | 1.13× |
| 4× | 63.4 | 2.33× |