← Back to community benchmarks
Qwen3.5-27B-Architect-Polaris-Heretic-mxfp4
Performance
4k
tokens
122.2
PP tok/s
15.0
TG tok/s
33510
TTFT (ms)
16.2
Peak mem (GB)
Hardware
Chip
M4 Pro (20c)
Memory
48 GB
GPU Cores
20
Software
oMLX
v0.2.8
macOS
macOS 26.2
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 122.0 | 15.2 | 15.9 GB | view |
| 4k | 122.2 | 15.0 | 16.2 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 15.2 | 1.00× |
| 2× | 30.0 | 1.97× |
| 4× | 36.8 | 2.42× |
| 8× | 71.6 | 4.71× |