← Back to community benchmarks
Qwen3.5-122B-A10B-oQ6
Performance
32k
tokens
358.0
PP tok/s
22.0
TG tok/s
91524
TTFT (ms)
100.5
Peak mem (GB)
Hardware
Chip
M3 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.4
macOS
macOS 26.4
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 399.2 | 33.3 | 95.0 GB | view |
| 4k | 473.4 | 31.3 | 96.3 GB | view |
| 8k | 470.4 | 30.9 | 96.9 GB | view |
| 16k | 435.2 | 28.3 | 98.1 GB | view |
| 32k | 358.0 | 22.0 | 100.5 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 33.3 | 1.00× |
| 2× | 44.7 | 1.34× |
| 4× | 58.4 | 1.75× |