← Back to community benchmarks
Qwen3.6-35B-A3BoQ8
Performance
16k
tokens
909.5
PP tok/s
53.5
TG tok/s
18014
TTFT (ms)
37.1
Peak mem (GB)
Hardware
Chip
M2 Max (38c)
Memory
96 GB
GPU Cores
38
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 | 931.2 | 69.0 | 35.4 GB | view |
| 4k | 1,148 | 64.5 | 36.2 GB | view |
| 8k | 953.8 | 63.2 | 36.5 GB | view |
| 16k | 909.5 | 53.5 | 37.1 GB | current |
| 32k | 789.7 | 43.6 | 38.5 GB | view |
| 64k | 611.5 | 32.7 | 41.2 GB | view |
| 128k | 412.2 | 18.1 | 46.5 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 69.0 | 1.00× |
| 2× | 102.0 | 1.48× |
| 4× | 115.7 | 1.68× |
| 8× | 159.6 | 2.31× |