← Back to community benchmarks
Qwen3.6-27BoQ4
Performance
32k
tokens
313.3
PP tok/s
24.2
TG tok/s
104592
TTFT (ms)
22.4
Peak mem (GB)
Hardware
Chip
M3 Ultra (60c)
Memory
96 GB
GPU Cores
60
Software
oMLX
v0.3.6
macOS
macOS 26.4.1
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 322.8 | 38.1 | 16.6 GB | view |
| 8k | 344.1 | 34.7 | 18.7 GB | view |
| 32k | 313.3 | 24.2 | 22.4 GB | current |
| 64k | 272.8 | 16.1 | 27.5 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 38.1 | 1.00× |
| 2× | 61.8 | 1.62× |
| 4× | 70.8 | 1.86× |
| 8× | 79.3 | 2.08× |