← Back to community benchmarks
Qwen3.5-122B-A10B
Performance
64k
tokens
465.3
PP tok/s
29.7
TG tok/s
140850
TTFT (ms)
77.0
Peak mem (GB)
Hardware
Chip
M3 Ultra (60c)
Memory
256 GB
GPU Cores
60
Software
oMLX
v0.2.17
macOS
macOS 26.3
Context
65,536
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 711.1 | 53.5 | 66.4 GB | view |
| 4k | 757.0 | 50.9 | 67.9 GB | view |
| 8k | 740.4 | 48.4 | 68.5 GB | view |
| 16k | 693.4 | 44.3 | 69.7 GB | view |
| 32k | 606.9 | 38.1 | 72.1 GB | view |
| 64k | 465.3 | 29.7 | 77.0 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 53.5 | 1.00× |
| 2× | 85.7 | 1.60× |
| 4× | 122.5 | 2.29× |
| 8× | 165.7 | 3.10× |