Qwen

by Alibaba

About the model:

Reasoning model with claims of comparable performance to DeepSeek-R1.

Model type:

Text

Model performance class:

Latest

Link to the model:

https://huggingface.co/Qwen/Qwen3-235B-A22B

Base models:

Qwen3-235B-A22B-Base

End model:

Qwen3-235B-A22B

End model license:

Apache-2.0

About the organisation:

Alibaba Cloud, a subdivision of the Chinese multinational technology company Alibaba.

Link to the organisation:

https://www.alibabacloud.com/

Model release date:

April 2025

Availability

Base Model Data

Pretraining data not specified or documented.

End User Model Data

Data not specified or documented.

Base Model Weights

No base model made available.

End User Model Weights

Model weights made available on HuggingFace.

https://huggingface.co/Qwen/Qwen3-235B-A22B

Training Code

Repository provides sparse source code and some examples for SFT

https://github.com/QwenLM/Qwen3

Documentation

Code Documentation

Repository is fairly well-documented.

https://github.com/QwenLM/Qwen3

Hardware Architecture

No description of hardware architecture found.

Preprint

Preprint published on arXiv.

https://arxiv.org/abs/2505.09388

Paper

No peer-reviewed paper found.

Modelcard

Model card primarily contains usage instructions.

https://huggingface.co/Qwen/Qwen3-235B-A22B

Datasheet

No datasheet found.

Access

Package

Model available on Ollama.

https://ollama.com/library/qwen3

API and Meta Prompts

Model available through HuggingFace space.

https://huggingface.co/spaces/Qwen/Qwen3-Demo

Licenses

Model released under Apache-2.0, an OSI-approved license.

https://qwenlm.github.io/blog/qwen3/

Is this information not up to date?

Contribute here ->

Supported by the Centre for Language Studies and the Dutch Research Council. Website design & development © 2024 by BSTN. This version of the index generated 10 March 2026, website content last updated 11 March 2026.