← Back to community benchmarks
Qwen3.5-122B-A10B
Performance
8k
tokens
520.7
PP tok/s
43.0
TG tok/s
15733
TTFT (ms)
67.0
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.2.11
macOS
macOS 26.3
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 513.1 | 49.9 | 66.3 GB | view |
| 8k | 520.7 | 43.0 | 67.0 GB | current |
| 16k | 436.7 | 34.2 | 67.7 GB | view |
| 64k | 245.5 | 15.0 | 71.9 GB | view |
| 195k | 71.3 | 4.0 | 83.4 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 49.9 | 1.00× |
| 2× | 32.8 | 0.66× |
| 4× | 38.5 | 0.77× |