← Back to community benchmarks
Qwen3.5-122B-A10B
Performance
16k
tokens
323.7
PP tok/s
32.0
TG tok/s
50613
TTFT (ms)
67.4
Peak mem (GB)
Hardware
Chip
M3 Max (30c)
Memory
96 GB
GPU Cores
30
Software
oMLX
v0.5.1
macOS
macOS 26.5.1
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 373.8 | 37.0 | 65.6 GB | view |
| 4k | 391.8 | 36.1 | 66.7 GB | view |
| 8k | 363.5 | 34.2 | 67.1 GB | view |
| 16k | 323.7 | 32.0 | 67.4 GB | current |
| 32k | 266.7 | 23.8 | 67.2 GB | view |
| 64k | 188.5 | 17.2 | 67.4 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 37.0 | 1.00× |
| 2× | 39.2 | 1.06× |
| 4× | 48.6 | 1.31× |
| 8× | 62.9 | 1.70× |