Qwen3-Coder 30B-A3B Instruct
Qwen3-Coder 30B-A3B Instruct is an open-weights language model from Qwen, released April 1, 2025. It has 31B parameters (mixture-of-experts, 3B active). The context window is 256K tokens. API price: $0.45 input / $2.25 output per million tokens.
Check it against your own PC →
Can you run Qwen3-Coder 30B-A3B Instruct locally?
The smallest common GPU it runs on is 6 GB of VRAM (with 32 GB of RAM): Q6_K, GPU + experts in RAM, roughly 19 tokens per second. Estimates assume 32 GB of DDR5 RAM, a 16K context and llama.cpp-style GGUF quantization. Enter your exact hardware in the live app.
| GPU memory | Verdict | Best quant | VRAM | RAM | tok/s |
|---|---|---|---|---|---|
| 6 GB VRAM | Runs · GPU + CPU experts | Q6_K (≈ lossless) | 5.2 GB | 22.0 GB | ~19 |
| 8 GB VRAM | Runs · GPU + CPU experts | Q6_K (≈ lossless) | 7.2 GB | 20.0 GB | ~22 |
| 12 GB VRAM | Runs · GPU + CPU experts | Q6_K (≈ lossless) | 11.2 GB | 16.0 GB | ~27 |
| 16 GB VRAM | Runs · GPU + CPU experts | Q6_K (≈ lossless) | 15.2 GB | 12.0 GB | ~34 |
| 24 GB VRAM | Runs on GPU | Q4_K_M (very good) | 20.9 GB | – | ~230 |
| 32 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 27.2 GB | – | ~300 |
Ready-made quantizations on Hugging Face: unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF · lmstudio-community/Qwen3-Coder-30B-A3B-Instruct-GGUF · ggml-org/Qwen3-Coder-30B-A3B-Instruct-GGUF · cyankiwi/Qwen3-Coder-30B-A3B-Instruct-AWQ-4bit · dark-side-of-the-code/Qwen3-Coder-30B-A3B-Instruct-AWQ · stelterlab/Qwen3-Coder-30B-A3B-Instruct-AWQ
Benchmark results
Epoch AI has not published results for Qwen3-Coder 30B-A3B Instruct yet. New models are usually evaluated within days to weeks.
Links
Hugging Face: Qwen/Qwen3-Coder-30B-A3B-Instruct
More from Qwen
Qwen3.8 Omni Flash · Qwen3.8 Max (0902) · Qwen3.8 Flash · Qwen3.8 Flash Next · Qwen 3.8 27B · Qwen3.8-2.4T-A95B · Qwen 3.8 Max · Qwen3.7 Flash