← Back to community benchmarks
Qwen3.5-122B-A10B
Performance
32k
tokens
768.2
PP tok/s
32.7
TG tok/s
42656
TTFT (ms)
128.1
Peak mem (GB)
Hardware
Chip
M3 Ultra (80c)
Memory
512 GB
GPU Cores
80
Software
oMLX
v0.3.0
macOS
macOS 26.3.1
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 756.4 | 43.6 | 122.5 GB | view |
| 4k | 923.1 | 41.8 | 123.9 GB | view |
| 8k | 924.7 | 40.2 | 124.5 GB | view |
| 16k | 877.8 | 37.3 | 125.7 GB | view |
| 32k | 768.2 | 32.7 | 128.1 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 43.6 | 1.00× |
| 2× | 66.6 | 1.53× |
| 4× | 131.3 | 3.01× |
| 8× | 263.3 | 6.04× |