← Back to community benchmarks
Qwen3.5-122B-A10B
Performance
1k
tokens
937.4
PP tok/s
49.2
TG tok/s
1092
TTFT (ms)
94.0
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.1
macOS
macOS 26.4
Context
1,024
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 937.4 | 49.2 | 94.0 GB | current |
| 4k | 1,371 | 47.5 | 95.5 GB | view |
| 8k | 1,354 | 45.5 | 96.1 GB | view |
| 16k | 1,225 | 41.3 | 97.3 GB | view |
| 32k | 998.8 | 34.3 | 99.7 GB | view |
| 64k | 525.5 | 21.0 | 104.5 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 49.2 | 1.00× |
| 2× | 51.1 | 1.04× |
| 4× | 62.5 | 1.27× |
| 8× | 76.1 | 1.55× |