← Back to community benchmarks
Qwen3.5-122B-A10B
Performance
16k
tokens
874.8
PP tok/s
35.3
TG tok/s
18728
TTFT (ms)
125.7
Peak mem (GB)
Hardware
Chip
M3 Ultra (80c)
Memory
512 GB
GPU Cores
80
Software
oMLX
v0.3.0
macOS
macOS 26.2
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 742.0 | 41.5 | 122.5 GB | view |
| 4k | 915.4 | 40.1 | 123.9 GB | view |
| 8k | 919.9 | 38.4 | 124.5 GB | view |
| 16k | 874.8 | 35.3 | 125.7 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 41.5 | 1.00× |
| 2× | 66.0 | 1.59× |
| 4× | 90.3 | 2.18× |
| 8× | 120.7 | 2.91× |