# Qwen3 Coder 30B-A3B Instruct: hardware requirements
> Qwen3 Coder 30B-A3B Instruct (Alibaba Qwen) is a 30.5B (3.3B active) mixture-of-experts model with a 256K-token context window, released under the Apache 2.0 license. It needs about 21GB of memory at Q4_K_M with 8K context.
Qwen3 Coder 30B-A3B is the compact member of Alibaba's 2025 coding line, built for agentic coding tools such as Qwen Code and Cline, with a native 256K context. It became the default local coding model of late 2025 because its 3.3B active parameters keep it fast on modest GPUs. Qwen3.6 35B-A3B has since overtaken it on agentic coding while adding vision and a thinking mode, so it is now superseded. It needs about 21GB at Q4_K_M with 8K context, fitting a 24GB GPU, and it runs well on 12GB to 16GB cards with experts offloaded to system RAM.

## Recommended hardware
- **Minimum GPU:** RTX 4000 Ada (20GB), Q3_K_M, ~84 tokens/sec
- **Best value:** RTX 3090 (24GB), Q4_K_M, ~135 tokens/sec
- **Fastest single GPU:** RTX 5090 (32GB), Q6_K, ~175 tokens/sec
- **Mac / unified memory:** Ryzen AI Max+ 395 (128GB), Q8_0, ~29 tokens/sec

## Memory by quantization
| Quant | Weights | Total @ 8K ctx | Total @ 32K ctx |
| --- | --- | --- | --- |
| Q2_K | 11.5GB | 13.3GB | 15.7GB |
| Q3_K_M | 15.0GB | 16.8GB | 19.3GB |
| Q4_K_M | 18.6GB | 20.6GB | 23.0GB |
| Q5_K_M | 21.9GB | 24.0GB | 26.4GB |
| Q6_K | 25.4GB | 27.5GB | 29.9GB |
| Q8_0 | 32.5GB | 34.9GB | 37.3GB |
| BF16 | 61.2GB | 64.4GB | 66.8GB |

## GPUs that run it fully in memory (fastest first)
| GPU | Memory | Quant | Speed | Price |
| --- | --- | --- | --- | --- |
| RTX 5090 | 32GB | Q6_K | ~175 tok/s | ~$4,300 |
| RTX PRO 6000 | 96GB | Q8_0 | ~150 tok/s | ~$16,000 |
| RTX 4090 | 24GB | Q4_K_M | ~150 tok/s | ~$2,500 used |
| RTX 3090 | 24GB | Q4_K_M | ~135 tok/s | ~$1,050 used |
| RTX PRO 5000 72GB | 72GB | Q8_0 | ~125 tok/s | ~$9,100 |
| RX 7900 XTX | 24GB | Q4_K_M | ~110 tok/s | ~$1,050 used |
| RTX 6000 Ada | 48GB | Q8_0 | ~99 tok/s | ~$4,750 used |
| V100 32GB | 32GB | Q6_K | ~92 tok/s | ~$700 used |
| RTX 4000 Ada | 20GB | Q3_K_M | ~84 tok/s | $1,250 MSRP |
| M5 Ultra 512GB | 512GB | Q8_0 | ~80 tok/s | n/a |
| M5 Ultra 256GB | 256GB | Q8_0 | ~80 tok/s | ~$9,499 |
| RTX A6000 | 48GB | Q8_0 | ~76 tok/s | ~$4,400 used |
| R9700 | 32GB | Q6_K | ~68 tok/s | ~$1,800 |
| MI50 32GB | 32GB | Q6_K | ~67 tok/s | ~$600 used |
| M3 Ultra 512GB | 512GB | Q8_0 | ~60 tok/s | $9,499 MSRP |
| Arc Pro B60 | 24GB | Q4_K_M | ~54 tok/s | ~$650 |
| M5 Max 128GB | 128GB | Q8_0 | ~54 tok/s | ~$5,100 |
| Arc Pro B70 | 32GB | Q6_K | ~54 tok/s | ~$1,300 |
| M4 Max 128GB | 128GB | Q8_0 | ~49 tok/s | $3,499 MSRP |
| Tesla P40 | 24GB | Q4_K_M | ~47 tok/s | ~$230 used |
| DGX Spark | 128GB | Q8_0 | ~31 tok/s | ~$6,950 |
| Ryzen AI Max+ PRO 495 | 192GB | Q8_0 | ~31 tok/s | ~$6,800 |
| Ryzen AI Max+ 395 | 128GB | Q8_0 | ~29 tok/s | ~$3,300 |

## How to run it
- Ollama: `ollama run qwen3-coder:30b`
- llama.cpp: `llama-server -hf unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF:Q4_K_M -c 8192 -ngl 99`
- Weights: https://huggingface.co/Qwen/Qwen3-Coder-30B-A3B-Instruct

## Strengths
- Tuned for repository-scale agentic coding with reliable tool calls
- Fast: 3.3B active parameters per token
- 256K native context for large codebases
- Apache 2.0 license

## Weaknesses
- No thinking mode, so it is weaker on hard algorithmic problems
- Superseded by Qwen3.6 35B-A3B for local agentic coding
- Narrow focus: general chat and writing lag general-purpose models
Full page: https://llamabuilds.ai/models/qwen3-coder-30b-a3b · Hardware guide: https://llamabuilds.ai/models/qwen3-coder-30b-a3b/requirements

---
Source: https://llamabuilds.ai/models/qwen3-coder-30b-a3b · Updated October 14, 2026 · Llama Builds. Speeds marked ~ are estimates from our bandwidth model (https://llamabuilds.ai/methodology); measured results are labeled as such.