← Back to community benchmarks
Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled-qx64-hi
Performance
32k
tokens
88.6
PP tok/s
8.2
TG tok/s
369769
TTFT (ms)
23.4
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
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 103.3 | 12.1 | 19.9 GB | view |
| 4k | 95.4 | 11.1 | 20.3 GB | view |
| 8k | 93.2 | 10.8 | 20.7 GB | view |
| 16k | 92.0 | 10.4 | 21.6 GB | view |
| 32k | 88.6 | 8.2 | 23.4 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 12.1 | 1.00× |
| 2× | 15.5 | 1.28× |
| 4× | 19.2 | 1.59× |
| 8× | 18.9 | 1.56× |