← Back to community benchmarks
Qwen3.5-27B
Performance
32k
tokens
170.6
PP tok/s
12.8
TG tok/s
192024
TTFT (ms)
32.6
Peak mem (GB)
Hardware
Chip
M2 Ultra (60c)
Memory
64 GB
GPU Cores
60
Software
oMLX
v0.2.11
macOS
macOS 26.3.1
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 179.8 | 20.7 | 29.1 GB | view |
| 4k | 183.1 | 20.1 | 29.5 GB | view |
| 8k | 181.7 | 18.4 | 29.9 GB | view |
| 16k | 178.0 | 15.4 | 30.8 GB | view |
| 32k | 170.6 | 12.8 | 32.6 GB | current |
| 64k | 155.4 | 9.5 | 36.2 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 20.7 | 1.00× |
| 2× | 33.6 | 1.62× |
| 4× | 42.7 | 2.06× |