← Back to community benchmarks
Qwen3.5-122B-A10B
Performance
16k
tokens
1,350
PP tok/s
49.1
TG tok/s
12135
TTFT (ms)
68.9
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.1
macOS
macOS 26.4
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,022 | 61.7 | 65.6 GB | view |
| 4k | 1,469 | 58.5 | 67.0 GB | view |
| 8k | 1,424 | 54.4 | 67.7 GB | view |
| 16k | 1,350 | 49.1 | 68.9 GB | current |
| 32k | 1,046 | 38.9 | 71.2 GB | view |
| 64k | 585.8 | 23.8 | 76.1 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 61.7 | 1.00× |
| 2× | 73.8 | 1.20× |
| 4× | 90.6 | 1.47× |
| 8× | 114.3 | 1.85× |