← Back to community benchmarks
Qwen3.5-9B
Performance
16k
tokens
310.1
PP tok/s
28.3
TG tok/s
52834
TTFT (ms)
11.7
Peak mem (GB)
Hardware
Chip
M1 Max (32c)
Memory
32 GB
GPU Cores
32
Software
oMLX
v0.2.6
macOS
macOS 26.2
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 314.9 | 35.4 | 10.7 GB | view |
| 4k | 318.6 | 34.5 | 11.0 GB | view |
| 8k | 316.4 | 31.9 | 11.2 GB | view |
| 16k | 310.1 | 28.3 | 11.7 GB | current |
| 32k | 288.6 | 22.5 | 12.7 GB | view |
| 64k | 220.1 | 15.6 | 14.8 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 35.4 | 1.00× |
| 2× | 43.1 | 1.22× |
| 4× | 51.4 | 1.45× |
| 8× | 48.5 | 1.37× |