← Back to community benchmarks
Qwen3.6-27B-oQ8-mtp
Performance
16k
tokens
168.1
PP tok/s
14.3
TG tok/s
97478
TTFT (ms)
32.8
Peak mem (GB)
Hardware
Chip
M3 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.9.dev2
macOS
macOS 26.3.1
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 200.7 | 20.0 | 28.8 GB | view |
| 4k | 193.7 | 18.9 | 30.3 GB | view |
| 8k | 165.4 | 17.2 | 31.3 GB | view |
| 16k | 168.1 | 14.3 | 32.8 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 20.0 | 1.00× |
| 2× | 18.1 | 0.91× |
| 4× | 24.7 | 1.23× |
| 8× | 30.5 | 1.52× |