← Back to community benchmarks
Qwen3-Coder-Next
Performance
32k
tokens
1,262
PP tok/s
45.8
TG tok/s
25975
TTFT (ms)
81.8
Peak mem (GB)
Hardware
Chip
M3 Ultra (60c)
Memory
256 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 | 1,236 | 63.9 | 80.1 GB | view |
| 4k | 1,509 | 61.9 | 80.3 GB | view |
| 8k | 1,521 | 56.5 | 80.5 GB | view |
| 16k | 1,434 | 52.0 | 80.9 GB | view |
| 32k | 1,262 | 45.8 | 81.8 GB | current |
| 64k | 994.8 | 34.7 | 83.6 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 63.9 | 1.00× |
| 2× | 106.5 | 1.67× |
| 4× | 166.3 | 2.60× |
| 8× | 239.1 | 3.74× |