← Back to community benchmarks
Qwen3.6-27B-Omnimerge-v4-oQ4-mtp
Performance
32k
tokens
228.1
PP tok/s
32.8
TG tok/s
143663
TTFT (ms)
23.7
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.9.dev2
macOS
macOS 26.5
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 241.7 | 41.4 | 16.8 GB | view |
| 4k | 250.0 | 39.4 | 18.2 GB | view |
| 8k | 248.2 | 38.2 | 19.2 GB | view |
| 16k | 241.6 | 34.6 | 20.7 GB | view |
| 32k | 228.1 | 32.8 | 23.7 GB | current |
| 64k | 202.8 | 26.1 | 29.7 GB | view |
| 128k | 163.8 | 20.4 | 42.1 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 41.4 | 1.00× |
| 2× | 50.4 | 1.22× |
| 4× | 57.0 | 1.38× |