← Back to community benchmarks
Qwen3.6-35B-A3B
Performance
64k
tokens
839.4
PP tok/s
27.6
TG tok/s
78079
TTFT (ms)
41.2
Peak mem (GB)
Hardware
Chip
M3 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.6
macOS
macOS 15.7.5
Context
65,536
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,108 | 70.0 | 35.4 GB | view |
| 4k | 1,390 | 67.2 | 36.2 GB | view |
| 8k | 1,390 | 62.1 | 36.5 GB | view |
| 16k | 1,305 | 56.0 | 37.1 GB | view |
| 32k | 1,071 | 41.3 | 38.5 GB | view |
| 64k | 839.4 | 27.6 | 41.2 GB | current |
| 128k | 492.2 | 11.3 | 46.5 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 70.0 | 1.00× |
| 2× | 101.3 | 1.45× |
| 4× | 127.2 | 1.82× |
| 8× | 161.6 | 2.31× |