← Back to community benchmarks
Qwen3.6-35B-A3B-AntiLoop
Performance
64k
tokens
569.6
PP tok/s
48.1
TG tok/s
115057
TTFT (ms)
23.2
Peak mem (GB)
Hardware
Chip
M2 Ultra (60c)
Memory
128 GB
GPU Cores
60
Software
oMLX
v0.5.1
macOS
macOS 26.5
Context
65,536
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 987.7 | 63.7 | 20.5 GB | view |
| 4k | 1,169 | 62.3 | 21.2 GB | view |
| 8k | 1,148 | 61.0 | 21.4 GB | view |
| 16k | 999.6 | 58.3 | 21.8 GB | view |
| 32k | 790.0 | 53.8 | 22.7 GB | view |
| 64k | 569.6 | 48.1 | 23.2 GB | current |
| 128k | 362.9 | 39.3 | 25.5 GB | view |
| 195k | 264.0 | 33.3 | 28.4 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 63.7 | 1.00× |
| 2× | 109.9 | 1.73× |
| 4× | 174.1 | 2.73× |
| 8× | 251.5 | 3.95× |