← Back to community benchmarks
Qwen3.5-9B
Performance
64k
tokens
343.6
PP tok/s
14.8
TG tok/s
190735
TTFT (ms)
14.8
Peak mem (GB)
Hardware
Chip
M4 Pro (20c)
Memory
24 GB
GPU Cores
20
Software
oMLX
v0.2.7
macOS
macOS 26.3.1
Context
65,536
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 423.8 | 26.4 | 10.7 GB | view |
| 4k | 448.7 | 27.6 | 11.0 GB | view |
| 8k | 443.0 | 26.7 | 11.2 GB | view |
| 16k | 426.0 | 23.9 | 11.7 GB | view |
| 32k | 392.1 | 19.2 | 12.7 GB | view |
| 64k | 343.6 | 14.8 | 14.8 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 26.4 | 1.00× |
| 2× | 54.6 | 2.07× |
| 4× | 72.1 | 2.73× |
| 8× | 72.7 | 2.75× |