Alibaba
Qwen3-Coder 30B A3B Instruct
Qwen Coder family, 30.5B parameters, recommended as Q4_K_M for first-pass local usage in V1.
Params
30.5B
Context
256K
License
Apache 2.0
Best runtime
Ollama
Recommended hardware
First-pass fit across priority GPUs
| Hardware | Fit | Decode | Safe ctx |
|---|---|---|---|
| NVIDIA A10 24GB | Tight fit | 39.7 tok/s | 17K |
| NVIDIA A100 40GB | Runs well | 142.9 tok/s | 27K |
| NVIDIA A100 80GB | Runs well | 187.4 tok/s | 46K |
| NVIDIA A16 64GB | Runs well | 55.2 tok/s | 39K |