I find this add-on extremely useful—far more so than the Be My Eyes Windows app. If you’re able to get access to a Google Gemini Pro Vision API key, you’ll be amazed at how quickly and accurately it can describe visual content.
You can request a description of the currently focused object, which helps avoid unnecessary UI elements and keeps the output relevant. Alternatively, you can describe a specific part of the screen, the entire screen, the currently focused file, or even an image from the clipboard. There's also an option to take pictures using your device’s built-in camera, which is especially helpful for framing yourself correctly before recording content or joining an online meeting.
Addon Link: https://github.com/cartertemm/AI-content-describer
By SeasonKing, 12 June, 2025
Forum
Windows
Comments
Thanks for reminding me about this
Hi,
I'd forgotten about this add-on. I've just set it up, both with my ChatGPT API key, and my Gemini one that I've just created. I'll have to try it out when I get something I need to describe.
Image Descriptions
I have enabled the addon for the AI description and gave it a go by pressin the NVDA, shift and I key and selected focus from the list. Nothing happens and I've even enabled the Chat GPT addon and the shortcut for that doesn't even work. These addons simply don't work and I don't understand what the hipe is all about.
a new NVDA update might be the reason.
aThe addon isn't compattable at the moment but you should be able to pass that by going to the addon store, finding the addon you've installed and pressing your applocations key, the narro down to something like enable addon, press enter on that, y for yes, y for yes again and NVDA should restart.
Then you should be good to go.
I installed the addon and it worked out of the box, so if that isnt' the reason; I don't know what's going on.
This works for me
Hi,
This worked for me, even with the 2025.1 update. I just enabled the add-on again, even though it was incompatible. I've downgraded now to NVDA version 2024.2 and it's still working fine. I've been using it a lot with Gemini's pro vision model, managed to add the API key no problems. And I've just tried it with the GPT pollinations thing, and it works fine. I've got the following NVDA settings all set to on. Carret moves review cursor, focus moves navigator object and automatically set system focus to focusable elements. I created a short cut for describing the current item which isn't available by default, and it describes what's under the NVDA review cursor now. And one of the options listed above was turned off, so that's why I couldn't get the description I wanted at first. Try using the 'entire screen' command to see what happens. If the image description cuts off, the maximum token amount might be too short. For Pollinations it's 300 by default, but you can change it in the 'manage models' dialog. The add-on is supposed to beep when it detects and starts processing the description.
@Saqib
Man, you are missing out. It's really amazing. Hope you manage to get it working with above responces. Default polinations AI is a bit slow, but still faster than BME. Google Gemini Vision Pro key changes the game. Literally under 4 seconds descriptions are spoken out, and that too quite consistantly. Plus, you can get it to describe only the focused object, unlike BME where it takes screenshot of entire screen.
P.S. Responce speed may be dependent on your internet speed etc.
Sounds like this might be a nice alternative…
Especially since people are having problems even logging onto the BME PC application. 🫤👎
Shush!
Shoooooooooosh! They will here you. They own this place.
Jokes aside, I am sure BME is working on enhancements of it's own to make their product unique and stand out from others. They were the earliest to the party and now, It's just that other people in the game have caught on. Nice! Some competition, keeps everyone on their tows, promotes innovation.