← Back to community benchmarks
Qwen3.6-35B-A3B-oQ8e
Performance
16k
tokens
843.6
PP tok/s
48.7
TG tok/s
19420
TTFT (ms)
37.1
Peak mem (GB)
Hardware
Chip
M1 Ultra (48c)
Memory
64 GB
GPU Cores
48
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 | 684.1 | 58.0 | 35.4 GB | view |
| 4k | 862.4 | 56.3 | 36.2 GB | view |
| 8k | 868.1 | 53.8 | 36.5 GB | view |
| 16k | 843.6 | 48.7 | 37.1 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 58.0 | 1.00× |
| 2× | 98.5 | 1.70× |
| 4× | 137.2 | 2.37× |
| 8× | 193.3 | 3.33× |