Qwen is the family of large language models built by Alibaba Cloud, with many sizes and versions released with open weights.
In plain words
Qwen is not one model but a series. The first release in 2023 included base models, chat models and specialist versions for code and maths. Later generations kept the name with a version number, such as Qwen2.5 and Qwen3. Because many Qwen models, including all of Qwen3, are released under the open Apache 2.0 licence, developers can download them and fine-tune them for their own needs.
A quick example
Imagine a start-up that needs a chatbot in several languages but cannot send customer data to an outside API. It could download a small Qwen model, fine-tune it on its own data and run it on its own server.
Where you'll meet it
- Open-weight chat, coding and reasoning models.
- A starting point for fine-tuning your own model.
Related
Sources
- DocsQwen, Hugging Face · read 29 Sept 2026
- PaperQwen Technical Report, Bai et al. (arXiv) · read 29 Sept 2026
- PaperQwen3 Technical Report, Qwen Team (arXiv) · read 29 Sept 2026
Checked by WhatIsAI against the sources above · 29 Sept 2026. Spotted an error? Report it.