← Back to community benchmarks
Qwen3.5-9B
Performance
1k
tokens
748.6
PP tok/s
69.2
TG tok/s
1368
TTFT (ms)
6.7
Peak mem (GB)
Hardware
Chip
M3 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.2.6
macOS
macOS 26.3
Context
1,024
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 748.6 | 69.2 | 6.7 GB | current |
| 4k | 772.4 | 65.6 | 7.0 GB | view |
| 8k | 759.7 | 62.6 | 7.1 GB | view |
| 16k | 686.5 | 57.8 | 7.6 GB | view |
| 32k | 634.3 | 47.3 | 8.7 GB | view |
| 64k | 490.2 | 22.5 | 10.8 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 69.2 | 1.00× |
| 2× | 63.4 | 0.92× |
| 4× | 89.8 | 1.30× |
| 8× | 103.3 | 1.49× |