← Back to community benchmarks
Qwen3.5-122B-A10B-Text-mxfp4
Performance
32k
tokens
439.2
PP tok/s
28.8
TG tok/s
74615
TTFT (ms)
64.7
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.2.6
macOS
macOS 26.3
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 469.3 | 56.4 | 62.0 GB | view |
| 4k | 552.8 | 54.4 | 62.3 GB | view |
| 8k | 538.9 | 47.6 | 62.6 GB | view |
| 16k | 485.0 | 40.1 | 63.3 GB | view |
| 32k | 439.2 | 28.8 | 64.7 GB | current |
| 64k | 358.0 | 19.2 | 67.5 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 56.4 | 1.00× |
| 2× | 88.7 | 1.57× |