← Back to community benchmarks
Qwen3.5-122B-A10B
Performance
8k
tokens
920.4
PP tok/s
47.4
TG tok/s
8901
TTFT (ms)
67.0
Peak mem (GB)
Hardware
Chip
M3 Ultra (80c)
Memory
512 GB
GPU Cores
80
Software
oMLX
v0.2.6
macOS
macOS 15.5
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 897.8 | 52.3 | 66.3 GB | view |
| 4k | 935.9 | 47.9 | 66.7 GB | view |
| 8k | 920.4 | 47.4 | 67.0 GB | current |
| 16k | 862.3 | 43.1 | 67.7 GB | view |
| 32k | 746.1 | 36.4 | 69.1 GB | view |
| 64k | 570.2 | 27.7 | 71.9 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 52.3 | 1.00× |
| 2× | 90.5 | 1.73× |
| 4× | 137.9 | 2.64× |
| 8× | 202.0 | 3.86× |