← Back to community benchmarks
Qwen3.5-122B-A10B
Performance
64k
tokens
148.2
PP tok/s
9.3
TG tok/s
442091
TTFT (ms)
71.9
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.2.8
macOS
macOS 26.3.1
Context
65,536
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 267.7 | 28.8 | 66.3 GB | view |
| 4k | 280.8 | 26.9 | 66.7 GB | view |
| 8k | 274.1 | 24.4 | 67.0 GB | view |
| 16k | 254.5 | 20.2 | 67.7 GB | view |
| 32k | 210.6 | 14.1 | 69.1 GB | view |
| 64k | 148.2 | 9.3 | 71.9 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 28.8 | 1.00× |
| 2× | 33.2 | 1.15× |
| 4× | 39.3 | 1.36× |
| 8× | 51.6 | 1.79× |