unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF

In plain English

Good for writing and refactoring code across common languages and explaining unfamiliar codebases. Still makes mistakes on edge cases, so treat its output as a draft that needs review, not a ready merge.

A 30 billion parameter mixture of experts model from Qwen's Coder line, quantised to GGUF by the community publisher unsloth for local use with llama.cpp style runners.

Required documentation

Complete

Intended use

Intended for code writing, refactoring, and codebase explanation with a human reviewing changes before merge. Not intended for fully autonomous commits.

Training data provenance

A GGUF quantisation of Qwen's Coder model line, trained on code and general text; the quantisation itself was produced by the community publisher unsloth, not Qwen directly.

Licence & terms

Apache-2.0, inherited from the base Qwen3 Coder release.

Reviews4

Write a review

MRMaya Rehanirep 4.8Productionb17cb02

7h ago

Accuracy4.4

Fine-tune4.0

Documentation3.8

Speed3.2

TNTomasz Nowakrep 4.5Productionb17cb02

7h ago

Accuracy4.6

Fine-tune3.8

Documentation4.0

Speed3.0

PSPriya Shahrep 4.6Fine-tunedb17cb02

7h ago

Accuracy4.5

Fine-tune4.1

Documentation3.6

Speed3.3

ACAlex Chenrep 4.4Prototypeb17cb02

7h ago

Accuracy4.3

Fine-tune3.9

Documentation3.9

Speed3.1

Discussion0

Posting as a signed-in userSign in to post

No questions yet. Be the first to ask one.

Structured rating

3.8

from 4 reviews of b17cb02

Only 4 reviews on b17cb02 so far. Earlier versions carry more, but they are not this build.

Accuracy4.5

Fine-tune3.9

Documentation3.8

Speed3.1

Run it now

ollama run unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF

Will it fit?

Build	File	VRAM
Q4_K_M	18.6 GB	20.5 GB
Q5_K_M	21.7 GB	23.5 GB

Set a hardware profile in the browse sidebar to see which builds fit.

Inference providers

No hosted providers listed yet.

No cloud required. Run it with

Ollamallama.cppLM StudioDocker Model Runner

Model tree

Base modelQwen/Qwen3-Coder-30B-A3B-Instruct

Adapters0

Finetunes0

Quantizations0

Merges0

At a glance

Parameters—

Licenceapache-2.0

Downloads12.8M

Reviews4

Docs completenessComplete