← Back to community benchmarks
Qwen3.5-35B-A3B
Performance
64k
tokens
1,225
PP tok/s
25.7
TG tok/s
53506
TTFT (ms)
23.3
Peak mem (GB)
Hardware
Chip
M3 Ultra (60c)
Memory
96 GB
GPU Cores
60
Software
oMLX
v0.2.6
macOS
macOS 26.3.1
Context
65,536
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,542 | 98.0 | 20.0 GB | view |
| 4k | 2,026 | 94.3 | 20.2 GB | view |
| 8k | 2,106 | 88.3 | 20.4 GB | view |
| 16k | 1,943 | 79.5 | 20.8 GB | view |
| 32k | 1,680 | 67.3 | 21.6 GB | view |
| 64k | 1,225 | 25.7 | 23.3 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 98.0 | 1.00× |
| 2× | 89.4 | 0.91× |
| 4× | 254.6 | 2.60× |