← Back to community benchmarks
Qwen3.6-27B-oQ4
Performance
1k
tokens
189.3
PP tok/s
21.1
TG tok/s
5408
TTFT (ms)
16.4
Peak mem (GB)
Hardware
Chip
M2 Max (38c)
Memory
96 GB
GPU Cores
38
Software
oMLX
v0.3.8rc1
macOS
macOS 26.4.1
Context
1,024
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 189.3 | 21.1 | 16.4 GB | current |
| 4k | 189.1 | 20.1 | 17.9 GB | view |
| 8k | 183.2 | 19.6 | 18.5 GB | view |
| 16k | 182.7 | 18.2 | 19.7 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 21.1 | 1.00× |
| 2× | 30.1 | 1.43× |
| 4× | 32.6 | 1.55× |