← Back to community benchmarks
Qwen3.6-35B-A3B-oQ4
Performance
16k
tokens
577.6
PP tok/s
44.2
TG tok/s
28366
TTFT (ms)
21.7
Peak mem (GB)
Hardware
Chip
M1 Max (32c)
Memory
64 GB
GPU Cores
32
Software
oMLX
v0.3.6
macOS
macOS 26.4.1
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 499.4 | 64.1 | 19.9 GB | view |
| 4k | 600.1 | 59.9 | 20.7 GB | view |
| 8k | 600.4 | 56.0 | 21.1 GB | view |
| 16k | 577.6 | 44.2 | 21.7 GB | current |
| 32k | 527.1 | 33.4 | 23.0 GB | view |
| 64k | 400.4 | 21.2 | 25.7 GB | view |
| 128k | 259.3 | 12.6 | 31.1 GB | view |
| 195k | 184.9 | 9.2 | 36.7 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 64.1 | 1.00× |
| 2× | 86.1 | 1.34× |
| 4× | 107.7 | 1.68× |
| 8× | 122.5 | 1.91× |