Alibaba releases open AI model for images

Alibaba has released a new AI model called Qwen-Image-2.1. The model can both analyze images and create new images from a description.

The model is built on 7 billion so-called parameters. Parameters are the settings the model uses to learn and recognize patterns; the more parameters, the more the model can typically handle, but it also requires more computing power to run.

Qwen-Image-2.1 is an open model. This means the code behind the model is publicly available, so other companies and developers can freely use, adapt and build on it, unlike closed models, where only the company itself has access to the underlying building blocks.

For ordinary Danes, the release could mean that more apps and services will be able to offer image analysis and generation in the future, because developers can adopt the model more cheaply and easily.

What this means for you

If you are a developer or work with IT at your company, you can download and try the model yourself without paying for a license. Most other people will only notice it when apps and programs they already use get new image features built in. If you are just an ordinary user, you don't need to do anything now; keep an eye out for whether your favorite apps start offering smarter image tools.