Qwen3 Coder, An Open-Weight Agentic Coding Model

There’s been a slew of Qwen releases recently. Among them is Qwen 3 Coder, an agentic Mixture of Experts (MoE) model with 405B parameters and 35B active parameters, designed for high-performance coding assistance and multi-turn tool use. The brief window (not even two weeks) between Kimi-K2 and Qwen 3 Coder’s debut demonstrates how aggressively teams are pushing specialized open-weight agentic coding models into the hands of developers. What makes this particular model stand out is its lower total parameter count (than say Kimi K2’s 1 trillion parameters) and impressive benchmark performance.

Qwen3 was released in May of this year, and in the conclusion of its technical report, they say “we will work on improving model architecture and training methods for the purposes of effective compression, scaling to extremely long contexts, etc. In addition, we plan to increase computational resources for reinforcement learning, with a particular emphasis on agent-based RL systems that learn from environmental feedback.”

Released in July, the updated Qwen3 model involves pretraining and reinforcement learning (RL) phases with a modified version of Group Relative Policy Optimization (GRPO) called Group Sequence Policy Optimization (GSPO) and a scalable system built to run 20 000 independent environments in parallel. We’re very excited (for the release of an updated technical report?) to learn more about the specifics.

• 405B parameter Mixture of Experts model with 35B active parameters
• 160 experts with 8 active per token
• 256K token context length extendable to 1M with YaRN
• High SWE-bench verified score on long horizon tasks (69.6 with 500 turns vs Claude-Sonnet-4 at 70.4% with 500 turns)
• Trained with Group Sequence Policy Optimization
• Smaller 30B A3B Instruct variant runs on a single H100 GPU
• Qwen Code CLI open-sourced as a fork of Gemini CLI

Here’s a high level overview to get you up to speed with Qwen 3 Coder’s internals.

On benchmarks, Qwen 3 Coder’s performance is remarkable. A score of 67.0% on SWE bench verified - which increases to 69.6% with 500 turns is incredible. The 500-turn result simulates a more realistic coding workflow - where the model can read feedback (like test failures), modify code, rerun tests, and repeat until the solution works.

This article will include implementation details for a smaller variant, Qwen3-Coder-30B-A3B-Instruct. For those curious about the name of this variant, there are 30 billion total parameters and 3 billion active parameters. The instruct indicates it’s an instruction-tuned variant of the base model.

So as we can see this particular model has slightly different specs, but can run on a single H100 GPU.

Step 1 : Set up a GPU Droplet

Begin by setting up a DigitalOcean GPU Droplet, select AI/ML and choose the NVIDIA H100 option.

Step 2: Web Console

Once your GPU Droplet finishes loading, you’ll be able to open up the Web Console.web console

Step 3: Install Dependendencies

apt install python3-pip
pip3 install transformers>=4.51.0

Step 4: Run the Model

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "Qwen/Qwen3-Coder-30B-A3B-Instruct"

# load the tokenizer and the model
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype="auto",
    device_map="auto"
)

# prepare the model input
prompt = "Write a quick sort algorithm."
messages = [
    {"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True,
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)

# conduct text completion
generated_ids = model.generate(
    **model_inputs,
    max_new_tokens=65536
)
output_ids = generated_ids[0][len(model_inputs.input_ids[0]):].tolist() 

content = tokenizer.decode(output_ids, skip_special_tokens=True)

print("content:", content)

Qwen Code is an open-source command-line interface for working with the Qwen3 Coder model on agentic coding tasks.
It’s a fork of the Gemini CLI, modified to work seamlessly with Qwen3’s capabilities.

We’ve included the steps to install the CLI, configure it, and run it with the Qwen3 Coder model.

Step 1: Install Node.js (Version 20 or Later)

Before you begin, make sure you have Node.js20+ installed on your device.
In your terminal:

node -v

Step 2: Install Qwen Code CLI

Once Node.js is ready, install Qwen Code globally:

npm install -g qwen-code

This makes the qwen-code command available from anywhere on your system.

Step 3: Get an API Key

Get an API key from openAI

export OPENAI_API_KEY="your_api_key_here"
export OPENAI_BASE_URL="https://dashscope-intl.aliyuncs.com/compatible-mode/v1"
export OPENAI_MODEL="qwen3-coder-plus"

Step 4: Vibe Code

Type qwen in your terminal and you’ll be able to vibe code.

For alternate ways to use Qwen3-Coder, check out the Qwen Coder blog post.

Here’s a notebook that may be of interest to those who want to improve their intuition around Qwen3’s underlying architecture.

Implement Qwen3 Mixture-of-Experts From Scratch by Sebastian Raschka: “this notebook runs Qwen3-Coder-30B-A3B-Instruct (aka Qwen3 Coder Flash) and requires 80 GB of VRAM (e.g., a single A100 or H100).”

We’re very excited to see the community play with these open-weight agentic coding models such as Qwen3-Coder, Kimi K2, Devstral, and integrate them in their workflows. What we’re most impressed about with Qwen3-Coder is its context window. At 246K tokens, extendable to a million, we’re excited to see how effective this model is in real-word software engineering use cases in comparison to alternative open-weight models. With its impressive context window, availability of accessible smaller variants with Qwen3-Coder-30B-A3B-Instruct, and the introduction of the Qwen Code CLI, this model is poised to empower developers with powerful, agentic coding assistance.