← Back to community benchmarks
Qwen3.5-122B-A10B-oQ4
Performance
16k
tokens
444.2
PP tok/s
31.6
TG tok/s
36887
TTFT (ms)
70.5
Peak mem (GB)
Hardware
Chip
M3 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.4
macOS
macOS 26.4
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 428.9 | 37.1 | 67.5 GB | view |
| 4k | 490.5 | 35.8 | 68.8 GB | view |
| 8k | 482.6 | 34.9 | 69.4 GB | view |
| 16k | 444.2 | 31.6 | 70.5 GB | current |
| 32k | 386.5 | 26.5 | 73.0 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 37.1 | 1.00× |
| 2× | 56.2 | 1.51× |
| 4× | 65.9 | 1.78× |