← Back to community benchmarks
Qwen3.5-122B-A10B-5bit
Performance
32k
tokens
470.0
PP tok/s
29.3
TG tok/s
69718
TTFT (ms)
86.3
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.2.24
macOS
macOS 26.4
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 540.3 | 42.4 | 80.6 GB | view |
| 4k | 578.0 | 40.6 | 82.1 GB | view |
| 8k | 566.3 | 38.8 | 82.7 GB | view |
| 16k | 531.5 | 35.3 | 83.9 GB | view |
| 32k | 470.0 | 29.3 | 86.3 GB | current |
| 64k | 371.0 | 17.4 | 91.2 GB | view |
| 128k | 260.5 | 10.8 | 100.9 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 42.4 | 1.00× |
| 2× | 61.1 | 1.44× |
| 4× | 74.3 | 1.75× |