← Back to community benchmarks
Qwen3-Coder-Next
Performance
64k
tokens
1,308
PP tok/s
43.6
TG tok/s
50117
TTFT (ms)
155.7
Peak mem (GB)
Hardware
Chip
M3 Ultra (80c)
Memory
512 GB
GPU Cores
80
Software
oMLX
v0.4.4
macOS
macOS 26.5.1
Context
65,536
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,323 | 51.9 | 149.6 GB | view |
| 4k | 2,006 | 51.8 | 150.5 GB | view |
| 64k | 1,308 | 43.6 | 155.7 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 51.9 | 1.00× |
| 2× | 62.8 | 1.21× |
| 4× | 103.0 | 1.98× |
| 8× | 63.2 | 1.22× |