← Back to community benchmarks
Qwen3.5-122B-A10B
Performance
8k
tokens
1,520
PP tok/s
59.4
TG tok/s
5391
TTFT (ms)
67.5
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.4
macOS
macOS 26.4
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,048 | 64.2 | 65.6 GB | view |
| 4k | 1,518 | 60.4 | 66.9 GB | view |
| 8k | 1,520 | 59.4 | 67.5 GB | current |
| 16k | 1,383 | 54.2 | 68.7 GB | view |
| 32k | 972.5 | 41.5 | 71.1 GB | view |
| 64k | 802.9 | 33.1 | 76.0 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 64.2 | 1.00× |
| 2× | 99.7 | 1.55× |
| 4× | 124.0 | 1.93× |
| 8× | 154.8 | 2.41× |