Quantization was done with an importance matrix that was trained for ~1M tokens (256 batches of 4096 tokens) of answers from the CodeFeedback-Filtered-Instruction dataset.

Fill-in-Middle tokens are automatically detected and supported as of commit 11ac980, see example.

Prompt template: ChatML

<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant

Compatibility

These quantised GGUFv3 files are compatible with llama.cpp from April 9th 2025 onwards, as of commit d3bd719

They are also compatible with many third party UIs and libraries provided they are built using a recent llama.cpp.

Explanation of quantisation methods

Click to see details

The new methods available are:

GGML_TYPE_IQ1_S - 1-bit quantization in super-blocks with an importance matrix applied, effectively using 1.56 bits per weight (bpw)
GGML_TYPE_IQ1_M - 1-bit quantization in super-blocks with an importance matrix applied, effectively using 1.75 bpw
GGML_TYPE_IQ2_XXS - 2-bit quantization in super-blocks with an importance matrix applied, effectively using 2.06 bpw
GGML_TYPE_IQ2_XS - 2-bit quantization in super-blocks with an importance matrix applied, effectively using 2.31 bpw
GGML_TYPE_IQ2_S - 2-bit quantization in super-blocks with an importance matrix applied, effectively using 2.5 bpw
GGML_TYPE_IQ2_M - 2-bit quantization in super-blocks with an importance matrix applied, effectively using 2.7 bpw
GGML_TYPE_IQ3_XXS - 3-bit quantization in super-blocks with an importance matrix applied, effectively using 3.06 bpw
GGML_TYPE_IQ3_XS - 3-bit quantization in super-blocks with an importance matrix applied, effectively using 3.3 bpw
GGML_TYPE_IQ3_S - 3-bit quantization in super-blocks with an importance matrix applied, effectively using 3.44 bpw
GGML_TYPE_IQ3_M - 3-bit quantization in super-blocks with an importance matrix applied, effectively using 3.66 bpw
GGML_TYPE_IQ4_XS - 4-bit quantization in super-blocks with an importance matrix applied, effectively using 4.25 bpw
GGML_TYPE_IQ4_NL - 4-bit non-linearly mapped quantization with an importance matrix applied, effectively using 4.5 bpw

Refer to the Provided Files table below to see what files use which methods, and how.

Provided files

Name	Quant method	Bits	Size	Max RAM required	Use case
Qwen3-Coder-30B-A3B-Instruct.IQ1_S.gguf (with YaRN)	IQ1_S	1	6.4 GB	6.7 GB	smallest, significant quality loss
Qwen3-Coder-30B-A3B-Instruct.IQ1_M.gguf (with YaRN)	IQ1_M	1	7.0 GB	7.3 GB	very small, significant quality loss
Qwen3-Coder-30B-A3B-Instruct.IQ2_XXS.gguf (with YaRN)	IQ2_XXS	2	8.0 GB	8.3 GB	very small, high quality loss
Qwen3-Coder-30B-A3B-Instruct.IQ2_XS.gguf (with YaRN)	IQ2_XS	2	8.8 GB	9.0 GB	very small, high quality loss
Qwen3-Coder-30B-A3B-Instruct.IQ2_S.gguf (with YaRN)	IQ2_S	2	9.0 GB	9.3 GB	small, substantial quality loss
Qwen3-Coder-30B-A3B-Instruct.IQ2_M.gguf (with YaRN)	IQ2_M	2	9.8 GB	10.0 GB	small, greater quality loss
Qwen3-Coder-30B-A3B-Instruct.IQ3_XXS.gguf (with YaRN)	IQ3_XXS	3	11.4 GB	11.7 GB	very small, high quality loss
Qwen3-Coder-30B-A3B-Instruct.IQ3_XS.gguf (with YaRN)	IQ3_XS	3	12.0 GB	12.3 GB	small, substantial quality loss
Qwen3-Coder-30B-A3B-Instruct.IQ3_S.gguf (with YaRN)	IQ3_S	3	12.7 GB	13.0 GB	small, greater quality loss
Qwen3-Coder-30B-A3B-Instruct.IQ3_M.gguf (with YaRN)	IQ3_M	3	12.9 GB	13.2 GB	medium, balanced quality
Qwen3-Coder-30B-A3B-Instruct.IQ4_XS.gguf (with YaRN)	IQ4_XS	4	15.5 GB	15.8 GB	small, marginal quality loss - recommended

Generated importance matrix file: Qwen3-Coder-30B-A3B-Instruct.imatrix.gguf

Note: the above RAM figures assume no GPU offloading with 4K context. If layers are offloaded to the GPU, this will reduce RAM usage and use VRAM instead.

Example `llama.cpp` command

Make sure you are using llama.cpp from commit d3bd719 or later.

./llama-cli -ngl 49 -m Qwen3-Coder-30B-A3B-Instruct.IQ4_XS.gguf --color -c 262144 --temp 0.7 --top-p 0.8 --top-k 20 --repeat-penalty 1.05 --jinja

Change -ngl 49 to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration.

Change -c 262144 to the desired sequence length.

If you are low on V/RAM try quantizing the K-cache with -ctk q8_0 or even -ctk q4_0 for big memory savings (depending on context size). There is a similar option for V-cache (-ctv), only available if you enable Flash Attention (-fa) as well.

For other parameters and how to use them, please refer to the llama.cpp documentation

How to run from Python code

You can use GGUF models from Python using the llama-cpp-python module.

How to load this model in Python code, using llama-cpp-python

For full documentation, please see: llama-cpp-python docs.

First install the package

Run one of the following commands, according to your system:

shell

# Prebuilt wheel with basic CPU support
pip install llama-cpp-python --extra-index-url https://abetlen.github.io/llama-cpp-python/whl/cpu
# Prebuilt wheel with NVidia CUDA acceleration
pip install llama-cpp-python --extra-index-url https://abetlen.github.io/llama-cpp-python/whl/cu121 (or cu122 etc.)
# Prebuilt wheel with Metal GPU acceleration
pip install llama-cpp-python --extra-index-url https://abetlen.github.io/llama-cpp-python/whl/metal
# Build base version with no GPU acceleration
pip install llama-cpp-python
# With NVidia CUDA acceleration
CMAKE_ARGS="-DGGML_CUDA=on" pip install llama-cpp-python
# Or with OpenBLAS acceleration
CMAKE_ARGS="-DGGML_BLAS=ON -DGGML_BLAS_VENDOR=OpenBLAS" pip install llama-cpp-python
# Or with AMD ROCm GPU acceleration (Linux only)
CMAKE_ARGS="-DGGML_HIPBLAS=on" pip install llama-cpp-python
# Or with Metal GPU acceleration for macOS systems only
CMAKE_ARGS="-DGGML_METAL=on" pip install llama-cpp-python
# Or with Vulkan acceleration
CMAKE_ARGS="-DGGML_VULKAN=on" pip install llama-cpp-python
# Or with SYCL acceleration
CMAKE_ARGS="-DGGML_SYCL=on -DCMAKE_C_COMPILER=icx -DCMAKE_CXX_COMPILER=icpx" pip install llama-cpp-python

# In windows, to set the variables CMAKE_ARGS in PowerShell, follow this format; eg for NVidia CUDA:
$env:CMAKE_ARGS = "-DGGML_CUDA=on"
pip install llama-cpp-python

Simple llama-cpp-python example code

python

from llama_cpp import Llama

# Chat Completion API

llm = Llama(model_path="./Qwen3-Coder-30B-A3B-Instruct.IQ4_XS.gguf", n_gpu_layers=49, n_ctx=262144)
print(llm.create_chat_completion(
    repeat_penalty = 1.05,
    messages = [
        {
            "role": "user",
            "content": "Pick a LeetCode challenge and solve it in Python."
        }
    ]
))

Simple llama-cpp-python example fill-in-middle code

python

from llama_cpp import Llama

# Completion API

prompt = "def add("
suffix = "\n    return sum\n\n"

llm = Llama(model_path="./Qwen3-Coder-30B-A3B-Instruct.IQ4_XS.gguf", n_gpu_layers=49, n_ctx=262144)
output = llm.create_completion(
    temperature = 0.0,
    repeat_penalty = 1.0,
    prompt = prompt,
    suffix = suffix
)

# Models sometimes repeat suffix in response, attempt to filter that
response = output["choices"][0]["text"]
response_stripped = response.rstrip()
unwanted_response_suffix = suffix.rstrip()
unwanted_response_length = len(unwanted_response_suffix)

filtered = False
if unwanted_response_suffix and response_stripped[-unwanted_response_length:] == unwanted_response_suffix:
    response = response_stripped[:-unwanted_response_length]
    filtered = True

print(f"Fill-in-Middle completion{' (filtered)' if filtered else ''}:\n\n{prompt}\033[32m{response}\033[{'33' if filtered else '0'}m{suffix}\033[0m")

Simple llama-cpp-python example function calling code

python

from llama_cpp import Llama

# Chat Completion API

grammar = LlamaGrammar.from_json_schema(json.dumps({
    "type": "array",
    "items": {
        "type": "object",
        "required": [ "name", "arguments" ],
        "properties": {
            "name": {
                "type": "string"
            },
            "arguments": {
                "type": "object"
            }
        }
    }
}))

llm = Llama(model_path="./Qwen3-Coder-30B-A3B-Instruct.IQ4_XS.gguf", n_gpu_layers=49, n_ctx=262144)
response = llm.create_chat_completion(
      temperature = 0.0,
      repeat_penalty = 1.05,
      messages = [
        {
          "role": "user",
          "content": "What's the weather like in Oslo and Stockholm?"
        }
      ],
      tools=[{
        "type": "function",
        "function": {
          "name": "get_current_weather",
          "description": "Get the current weather in a given location",
          "parameters": {
            "type": "object",
            "properties": {
              "location": {
                "type": "string",
                "description": "The city and state, e.g. San Francisco, CA"
              },
              "unit": {
                "type": "string",
                "enum": [ "celsius", "fahrenheit" ]
              }
            },
            "required": [ "location" ]
          }
        }
      }],
      grammar = grammar
)
print(json.loads(response["choices"][0]["text"]))

print(llm.create_chat_completion(
      temperature = 0.0,
      repeat_penalty = 1.05,
      messages = [
        {
          "role": "user",
          "content": "What's the weather like in Oslo?"
        },
        { # The tool_calls is from the response to the above with tool_choice active
          "role": "assistant",
          "content": None,
          "tool_calls": [
            {
              "id": "call__0_get_current_weather_cmpl-...",
              "type": "function",
              "function": {
                "name": "get_current_weather",
                "arguments": { "location": "Oslo, Norway" , "unit": "celsius" }
              }
            }
          ]
        },
        { # The tool_call_id is from tool_calls and content is the result from the function call you made
          "role": "tool",
          "content": "20",
          "tool_call_id": "call__0_get_current_weather_cmpl-..."
        }
      ],
      tools=[{
        "type": "function",
        "function": {
          "name": "get_current_weather",
          "description": "Get the current weather in a given location",
          "parameters": {
            "type": "object",
            "properties": {
              "location": {
                "type": "string",
                "description": "The city and state, e.g. San Francisco, CA"
              },
              "unit": {
                "type": "string",
                "enum": [ "celsius", "fahrenheit" ]
              }
            },
            "required": [ "location" ]
          }
        }
      }],
      #tool_choice={
      #  "type": "function",
      #  "function": {
      #    "name": "get_current_weather"
      #  }
      #}
))

Qwen3-Coder-30B-A3B-Instruct

Highlights

Qwen3-Coder is available in multiple sizes. Today, we're excited to introduce Qwen3-Coder-30B-A3B-Instruct. This streamlined model maintains impressive performance and efficiency, featuring the following key enhancements:

Significant Performance among open models on Agentic Coding, Agentic Browser-Use, and other foundational coding tasks.
Long-context Capabilities with native support for 256K tokens, extendable up to 1M tokens using Yarn, optimized for repository-scale understanding.
Agentic Coding supporting for most platform such as Qwen Code, CLINE, featuring a specially designed function call format.

Model Overview

Qwen3-Coder-30B-A3B-Instruct has the following features:

Type: Causal Language Models
Training Stage: Pretraining & Post-training
Number of Parameters: 30.5B in total and 3.3B activated
Number of Layers: 48
Number of Attention Heads (GQA): 32 for Q and 4 for KV
Number of Experts: 128
Number of Activated Experts: 8
Context Length: 262,144 natively.

NOTE: This model supports only non-thinking mode and does not generate <think></think> blocks in its output. Meanwhile, specifying enable_thinking=False is no longer required.

For more details, including benchmark evaluation, hardware requirements, and inference performance, please refer to our blog, GitHub, and Documentation.

Quickstart

We advise you to use the latest version of transformers.

With transformers<4.51.0, you will encounter the following error:

KeyError: 'qwen3_moe'

The following contains a code snippet illustrating how to use the model generate content based on given inputs.

python

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)

Note: If you encounter out-of-memory (OOM) issues, consider reducing the context length to a shorter value, such as 32,768.

For local use, applications such as Ollama, LMStudio, MLX-LM, llama.cpp, and KTransformers have also supported Qwen3.

Agentic Coding

Qwen3-Coder excels in tool calling capabilities.

You can simply define or use any tools as following example.

python

# Your tool implementation
def square_the_number(num: float) -> dict:
    return num ** 2

# Define Tools
tools=[
    {
        "type":"function",
        "function":{
            "name": "square_the_number",
            "description": "output the square of the number.",
            "parameters": {
                "type": "object",
                "required": ["input_num"],
                "properties": {
                    'input_num': {
                        'type': 'number', 
                        'description': 'input_num is a number that will be squared'
                        }
                },
            }
        }
    }
]

import OpenAI
# Define LLM
client = OpenAI(
    # Use a custom endpoint compatible with OpenAI API
    base_url='http://localhost:8000/v1',  # api_base
    api_key="EMPTY"
)
 
messages = [{'role': 'user', 'content': 'square the number 1024'}]

completion = client.chat.completions.create(
    messages=messages,
    model="Qwen3-Coder-30B-A3B-Instruct",
    max_tokens=65536,
    tools=tools,
)

print(completion.choice[0])

Best Practices

To achieve optimal performance, we recommend the following settings:

Sampling Parameters:
- We suggest using temperature=0.7, top_p=0.8, top_k=20, repetition_penalty=1.05.
Adequate Output Length: We recommend using an output length of 65,536 tokens for most queries, which is adequate for instruct models.

Citation

If you find our work helpful, feel free to give us a cite.

@misc{qwen3technicalreport,
      title={Qwen3 Technical Report}, 
      author={Qwen Team},
      year={2025},
      eprint={2505.09388},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2505.09388}, 
}

CISCai/Qwen3-Coder-30B-A3B-Instruct-SOTA-GGUF

Qwen3-Coder-30B-A3B-Instruct - SOTA GGUF

Description

Prompt template: ChatML

Compatibility

Explanation of quantisation methods

Provided files

Example llama.cpp command

How to run from Python code

How to load this model in Python code, using llama-cpp-python

First install the package

Simple llama-cpp-python example code

Simple llama-cpp-python example fill-in-middle code

Simple llama-cpp-python example function calling code

Qwen3-Coder-30B-A3B-Instruct

Highlights

Model Overview

Quickstart

Agentic Coding

Best Practices

Citation

CISCai/
Qwen3-Coder-30B-A3B-Instruct-SOTA-GGUF

Example `llama.cpp` command