← Back to community benchmarks
Qwen3.5-27B-Claude-4.6-Opus-Distilled
Performance
4k
tokens
417.6
PP tok/s
34.6
TG tok/s
9807
TTFT (ms)
16.1
Peak mem (GB)
Hardware
Chip
M3 Ultra (80c)
Memory
512 GB
GPU Cores
80
Software
oMLX
v0.2.12
macOS
macOS 15.6.1
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 388.1 | 36.3 | 15.8 GB | view |
| 4k | 417.6 | 34.6 | 16.1 GB | current |
| 8k | 416.7 | 32.7 | 16.6 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 36.3 | 1.00× |
| 2× | 65.4 | 1.80× |
| 4× | 89.3 | 2.46× |
| 8× | 102.9 | 2.83× |