← Back to community benchmarks
Qwen3.5-122B-A10B
Performance
4k
tokens
1,446
PP tok/s
62.7
TG tok/s
2833
TTFT (ms)
66.9
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.5.dev1
macOS
macOS 26.4.1
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 913.6 | 65.3 | 65.6 GB | view |
| 4k | 1,446 | 62.7 | 66.9 GB | current |
| 8k | 1,527 | 59.4 | 67.5 GB | view |
| 16k | 1,406 | 52.1 | 68.7 GB | view |
| 32k | 1,143 | 42.3 | 71.1 GB | view |
| 64k | 834.5 | 30.3 | 76.0 GB | view |
| 128k | 482.4 | 17.4 | 85.5 GB | view |
| 195k | 347.4 | 13.7 | 95.7 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 65.3 | 1.00× |
| 2× | 101.4 | 1.55× |
| 4× | 114.1 | 1.75× |