← Back to community benchmarks
Qwen3.6-27B-oQ4-fp16-mtp
Performance
16k
tokens
144.0
PP tok/s
16.4
TG tok/s
113780
TTFT (ms)
20.4
Peak mem (GB)
Hardware
Chip
M1 Max (32c)
Memory
64 GB
GPU Cores
32
Software
oMLX
v0.3.8
macOS
macOS 15.6.1
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 145.2 | 16.4 | 16.4 GB | view |
| 4k | 149.0 | 14.7 | 17.9 GB | view |
| 8k | 147.8 | 17.9 | 18.5 GB | view |
| 16k | 144.0 | 16.4 | 20.4 GB | current |
| 32k | 133.2 | 14.4 | 23.4 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 16.4 | 1.00× |
| 2× | 25.0 | 1.52× |
| 4× | 23.9 | 1.46× |