← Back to community benchmarks
Qwen3.5-122B-A10B
Performance
8k
tokens
408.3
PP tok/s
37.0
TG tok/s
20065
TTFT (ms)
67.0
Peak mem (GB)
Hardware
Chip
M3 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.2.6
macOS
macOS 26.3.1
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 465.1 | 45.4 | 66.3 GB | view |
| 4k | 476.3 | 41.9 | 66.7 GB | view |
| 8k | 408.3 | 37.0 | 67.0 GB | current |
| 16k | 366.9 | 29.0 | 67.7 GB | view |
| 32k | 288.6 | 16.3 | 69.1 GB | view |
| 64k | 135.5 | 14.0 | 71.9 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 45.4 | 1.00× |
| 2× | 47.9 | 1.06× |
| 4× | 61.0 | 1.34× |