← Back to community benchmarks
Qwen3.5-122B-A10B
Performance
16k
tokens
436.6
PP tok/s
35.1
TG tok/s
37530
TTFT (ms)
67.7
Peak mem (GB)
Hardware
Chip
M2 Ultra (76c)
Memory
192 GB
GPU Cores
76
Software
oMLX
v0.2.6
macOS
macOS 15.0
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 433.3 | 43.4 | 66.3 GB | view |
| 4k | 442.5 | 36.2 | 66.7 GB | view |
| 8k | 454.6 | 38.3 | 67.0 GB | view |
| 16k | 436.6 | 35.1 | 67.7 GB | current |
| 32k | 403.3 | 28.4 | 69.1 GB | view |
| 64k | 324.9 | 15.2 | 71.9 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 43.4 | 1.00× |
| 2× | 69.3 | 1.60× |
| 4× | 98.0 | 2.26× |
| 8× | 127.4 | 2.94× |