← Back to community benchmarks
Qwen3.5-122B-A10B-oQ4
Performance
16k
tokens
1,466
PP tok/s
15.1
TG tok/s
11175
TTFT (ms)
71.7
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.5.dev1
macOS
macOS 26.3.1
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 520.4 | 49.6 | 68.2 GB | view |
| 4k | 586.6 | 47.7 | 69.5 GB | view |
| 8k | 577.7 | 45.6 | 70.1 GB | view |
| 16k | 1,466 | 15.1 | 71.7 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 49.6 | 1.00× |
| 2× | 74.8 | 1.51× |
| 4× | 89.8 | 1.81× |
| 8× | 108.1 | 2.18× |