← Back to community benchmarks
Qwen3.5-122B-A10B
Performance
32k
tokens
776.9
PP tok/s
24.1
TG tok/s
42177
TTFT (ms)
235.6
Peak mem (GB)
Hardware
Chip
M3 Ultra (80c)
Memory
512 GB
GPU Cores
80
Software
oMLX
v0.2.19
macOS
macOS 26.3.1
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 771.1 | 29.4 | 229.9 GB | view |
| 4k | 956.0 | 28.6 | 231.4 GB | view |
| 8k | 949.5 | 27.8 | 232.0 GB | view |
| 16k | 894.8 | 26.4 | 233.2 GB | view |
| 32k | 776.9 | 24.1 | 235.6 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 29.4 | 1.00× |
| 2× | 32.7 | 1.11× |
| 4× | 51.9 | 1.77× |
| 8× | 42.7 | 1.45× |