← Back to community benchmarks
Qwen3.5-122B-A10B-oQ3.5
Performance
4k
tokens
724.8
PP tok/s
51.6
TG tok/s
5651
TTFT (ms)
59.6
Peak mem (GB)
Hardware
Chip
M3 Ultra (60c)
Memory
96 GB
GPU Cores
60
Software
oMLX
v0.3.2
macOS
macOS 26.4
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 558.4 | 52.0 | 58.1 GB | view |
| 4k | 724.8 | 51.6 | 59.6 GB | current |
| 8k | 732.3 | 49.6 | 60.2 GB | view |
| 16k | 691.3 | 47.4 | 61.4 GB | view |
| 32k | 607.4 | 42.2 | 63.8 GB | view |
| 64k | 467.0 | 35.4 | 68.6 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 52.0 | 1.00× |
| 2× | 80.0 | 1.54× |
| 4× | 103.2 | 1.98× |