← Back to community benchmarks
Qwen3.6-27B-oQ8-mtp
Performance
1k
tokens
293.4
PP tok/s
30.9
TG tok/s
3491
TTFT (ms)
28.8
Peak mem (GB)
Hardware
Chip
M3 Ultra (60c)
Memory
256 GB
GPU Cores
60
Software
oMLX
v0.4.0
macOS
macOS 26.5.1
Context
1,024
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 293.4 | 30.9 | 28.8 GB | current |
| 4k | 318.1 | 31.0 | 30.3 GB | view |
| 8k | 314.4 | 29.5 | 30.9 GB | view |
| 16k | 307.5 | 24.4 | 32.1 GB | view |
| 32k | 289.5 | 25.6 | 34.6 GB | view |
| 64k | 254.3 | 21.9 | 39.7 GB | view |
| 128k | 201.2 | 16.7 | 50.0 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 30.9 | 1.00× |
| 2× | 30.0 | 0.97× |
| 4× | 41.3 | 1.34× |