← Back to community benchmarks
Qwen3.5-122B-A10B
Performance
64k
tokens
237.6
PP tok/s
13.3
TG tok/s
275806
TTFT (ms)
71.9
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.2.7
macOS
macOS 15.7.4
Context
65,536
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 554.6 | 52.6 | 66.3 GB | view |
| 4k | 422.0 | 44.3 | 66.7 GB | view |
| 8k | 356.5 | 42.7 | 67.0 GB | view |
| 16k | 344.8 | 30.8 | 67.7 GB | view |
| 32k | 313.4 | 21.7 | 69.1 GB | view |
| 64k | 237.6 | 13.3 | 71.9 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 52.6 | 1.00× |
| 2× | 57.8 | 1.10× |
| 4× | 66.8 | 1.27× |
| 8× | 80.7 | 1.53× |