← Back to community benchmarks
Qwen3-Coder-Next
Performance
32k
tokens
737.4
PP tok/s
39.8
TG tok/s
44437
TTFT (ms)
44.7
Peak mem (GB)
Hardware
Chip
M2 Ultra (60c)
Memory
64 GB
GPU Cores
60
Software
oMLX
v0.2.6
macOS
macOS 26.3
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 718.1 | 61.8 | 43.0 GB | view |
| 4k | 827.3 | 58.9 | 43.2 GB | view |
| 8k | 828.5 | 54.8 | 43.4 GB | view |
| 16k | 801.8 | 47.6 | 43.9 GB | view |
| 32k | 737.4 | 39.8 | 44.7 GB | current |
| 64k | 617.2 | 26.5 | 46.5 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 61.8 | 1.00× |
| 2× | 99.8 | 1.61× |
| 4× | 147.0 | 2.38× |
| 8× | 200.1 | 3.24× |