← Back to community benchmarks
qwen3.6-27b-oQ8-mtp
Performance
16k
tokens
310.2
PP tok/s
28.3
TG tok/s
52826
TTFT (ms)
32.8
Peak mem (GB)
Hardware
Chip
M3 Ultra (60c)
Memory
256 GB
GPU Cores
60
Software
oMLX
v0.3.9.dev2
macOS
macOS 26.2
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 298.3 | 28.2 | 28.8 GB | view |
| 4k | 319.4 | 31.9 | 30.3 GB | view |
| 8k | 318.3 | 30.3 | 31.3 GB | view |
| 16k | 310.2 | 28.3 | 32.8 GB | current |
| 32k | 292.2 | 27.1 | 35.8 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 28.2 | 1.00× |
| 2× | 39.3 | 1.39× |
| 4× | 50.2 | 1.78× |