← Back to community benchmarks
Qwen3.6-27B-uncensored-heretic-v2-Native-MTP-Preserved-oQ4
Performance
16k
tokens
726.8
PP tok/s
37.4
TG tok/s
22541
TTFT (ms)
20.7
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.4.1
macOS
macOS 26.5.1
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 771.4 | 42.5 | 16.8 GB | view |
| 4k | 870.8 | 43.0 | 18.2 GB | view |
| 8k | 808.0 | 41.0 | 19.2 GB | view |
| 16k | 726.8 | 37.4 | 20.7 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 42.5 | 1.00× |
| 2× | 50.1 | 1.18× |
| 4× | 67.9 | 1.60× |