Please also update the following lines in changes.md:

You can now use NVDA+Windows+, to describe images you encounter.

• Press NVDA+Windows+, to get an AI generated image description. (Support image descriptions using local AI model #18475, @tianzeshi-study)

The following content also needs to be updated:

Once the Image Captioner is ready, press the default shortcut NVDA+Windows+, to recognize the image currently navigated by NVDA.

Please also update the following lines in changes.md:

You can now use NVDA+Windows+, to describe images you encounter.

• Press NVDA+Windows+, to get an AI generated image description. (Support image descriptions using local AI model #18475, @tianzeshi-study)

Thanks! I forgot it.

While at it, I think that UI and User Guide may be reworded for clarity to distinguish between image description, image captioner, model, etc. What is the difference, from a user's point of view, between captioner and model?

For example, there is nothing automatic in current image description feature:

When this checkbox is enabled, NVDA will enable automatic image descriptions.
I'd rather write something like:
When this checkbox is enabled, NVDA will load the model in memory, enabling the use of the image description command.

please note:

CI/CD / Check translator comments (pull_request)Failing after 1m

Regarding the wording (model / image captioner / image description), I have reviewed the code changed in this PR, but there is probably some other places to change to be consistent (GUI, script description, ...)