← Back to community benchmarks
unsloth--Qwen3.6-27B
Performance
195k
tokens
208.0
PP tok/s
8.9
TG tok/s
961535
TTFT (ms)
73.5
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.5.1
macOS
macOS 26.5.2
Context
200,000
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 833.4 | 12.5 | 33.0 GB | view |
| 4k | 381.9 | 12.6 | 35.6 GB | view |
| 8k | 398.1 | 12.2 | 36.5 GB | view |
| 16k | 457.4 | 14.1 | 38.0 GB | view |
| 32k | 450.9 | 13.5 | 41.3 GB | view |
| 64k | 368.5 | 12.2 | 47.3 GB | view |
| 128k | 259.2 | 10.1 | 59.6 GB | view |
| 195k | 208.0 | 8.9 | 73.5 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 12.5 | 1.00× |
| 2× | 16.5 | 1.32× |
| 4× | 29.6 | 2.37× |
| 8× | 35.0 | 2.80× |