← Back to community benchmarks
Qwen3.5-27B-Claude-4.6-OS-INSTRUCT-mxfp8
Performance
1k
tokens
114.0
PP tok/s
9.8
TG tok/s
8984
TTFT (ms)
28.4
Peak mem (GB)
Hardware
Chip
M2 Max (38c)
Memory
96 GB
GPU Cores
38
Software
oMLX
v0.2.6
macOS
macOS 26.3.1
Context
1,024
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 114.0 | 9.8 | 28.4 GB | current |
| 4k | 104.0 | 9.3 | 28.7 GB | view |
| 8k | 101.1 | 9.0 | 29.2 GB | view |
| 16k | 99.5 | 8.3 | 30.0 GB | view |
| 32k | 95.3 | 7.1 | 31.8 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 9.8 | 1.00× |
| 2× | 11.3 | 1.15× |
| 4× | 13.5 | 1.38× |
| 8× | 16.8 | 1.71× |