← Back to community benchmarks
Qwen3.5-27B
Performance
8k
tokens
318.8
PP tok/s
33.5
TG tok/s
25698
TTFT (ms)
17.4
Peak mem (GB)
Hardware
Chip
M3 Ultra (60c)
Memory
256 GB
GPU Cores
60
Software
oMLX
v0.2.9
macOS
macOS 26.3
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 320.7 | 37.7 | 16.6 GB | view |
| 4k | 324.7 | 35.6 | 17.0 GB | view |
| 8k | 318.8 | 33.5 | 17.4 GB | current |
| 16k | 310.2 | 30.2 | 18.3 GB | view |
| 32k | 291.7 | 24.5 | 20.1 GB | view |
| 64k | 259.2 | 19.0 | 23.6 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 37.7 | 1.00× |
| 2× | 60.7 | 1.61× |
| 4× | 71.0 | 1.88× |
| 8× | 77.4 | 2.05× |