← Back to community benchmarks
Qwen3.6-27B-MXFP4-CRACK-MTP
Performance
4k
tokens
171.7
PP tok/s
26.2
TG tok/s
23856
TTFT (ms)
19.7
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.4.1
macOS
macOS 26.4.1
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 218.5 | 23.1 | 16.9 GB | view |
| 4k | 171.7 | 26.2 | 19.7 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 23.1 | 1.00× |
| 2× | 39.4 | 1.71× |
| 4× | 47.3 | 2.05× |