← Back to community benchmarks
Qwen3.6-35B-A3B-oQ4
Performance
16k
tokens
1,123
PP tok/s
60.0
TG tok/s
14585
TTFT (ms)
21.7
Peak mem (GB)
Hardware
Chip
M2 Ultra (60c)
Memory
64 GB
GPU Cores
60
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 | 870.3 | 85.5 | 19.9 GB | view |
| 4k | 1,141 | 79.4 | 20.7 GB | view |
| 8k | 1,159 | 75.7 | 21.1 GB | view |
| 16k | 1,123 | 60.0 | 21.7 GB | current |
| 32k | 1,023 | 43.9 | 23.0 GB | view |
| 64k | 432.1 | 22.2 | 25.7 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 85.5 | 1.00× |
| 2× | 78.2 | 0.91× |
| 4× | 96.3 | 1.13× |
| 8× | 104.9 | 1.23× |