← Back to community benchmarks
qwen3.6-27b-oQ8-mtp
Performance
32k
tokens
292.0
PP tok/s
4.6
TG tok/s
112228
TTFT (ms)
40.5
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
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 298.9 | 32.3 | 28.8 GB | view |
| 4k | 319.4 | 32.2 | 30.3 GB | view |
| 8k | 318.3 | 31.2 | 31.3 GB | view |
| 16k | 310.1 | 28.4 | 32.8 GB | view |
| 32k | 292.0 | 4.6 | 40.5 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 32.3 | 1.00× |
| 2× | 14.2 | 0.44× |
| 4× | 25.7 | 0.80× |