← Back to community benchmarks
Qwen3.5-4B
Performance
64k
tokens
855.5
PP tok/s
50.5
TG tok/s
76602
TTFT (ms)
8.6
Peak mem (GB)
Hardware
Chip
M2 Ultra (60c)
Memory
64 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 | 969.4 | 129.0 | 4.0 GB | view |
| 4k | 1,112 | 123.1 | 4.2 GB | view |
| 8k | 1,132 | 113.4 | 4.6 GB | view |
| 16k | 1,087 | 98.1 | 5.3 GB | view |
| 32k | 1,002 | 76.4 | 6.9 GB | view |
| 64k | 855.5 | 50.5 | 8.6 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 129.0 | 1.00× |
| 2× | 181.6 | 1.41× |
| 4× | 220.5 | 1.71× |
| 8× | 249.8 | 1.94× |