← Back to community benchmarks
Qwen3.5-122B-A10B
Performance
16k
tokens
437.2
PP tok/s
28.8
TG tok/s
37472
TTFT (ms)
67.7
Peak mem (GB)
Hardware
Chip
M2 Ultra (76c)
Memory
192 GB
GPU Cores
76
Software
oMLX
v0.2.6
macOS
macOS 15.0
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 434.7 | 41.0 | 66.3 GB | view |
| 4k | 456.5 | 37.5 | 66.7 GB | view |
| 8k | 454.1 | 32.3 | 67.0 GB | view |
| 16k | 437.2 | 28.8 | 67.7 GB | current |
| 32k | 403.5 | 21.8 | 69.1 GB | view |
| 64k | 324.3 | 15.2 | 71.9 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 41.0 | 1.00× |
| 2× | 55.7 | 1.36× |
| 4× | 92.8 | 2.26× |
| 8× | 126.1 | 3.08× |