← Back to community benchmarks
Qwen3.6-27B-Opus-Distill-oQ8-mtp
Performance
8k
tokens
317.2
PP tok/s
30.2
TG tok/s
25822
TTFT (ms)
31.3
Peak mem (GB)
Hardware
Chip
M3 Ultra (60c)
Memory
256 GB
GPU Cores
60
Software
oMLX
v0.3.9rc1
macOS
macOS 26.2
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 298.6 | 29.2 | 28.8 GB | view |
| 4k | 318.6 | 31.7 | 30.3 GB | view |
| 8k | 317.2 | 30.2 | 31.3 GB | current |
| 16k | 309.3 | 29.3 | 32.8 GB | view |
| 32k | 291.0 | 2.3 | 61.6 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 29.2 | 1.00× |
| 2× | 19.4 | 0.66× |
| 4× | 44.5 | 1.52× |