AI content describer add-on no longer describing images, does anyone know of an alternative/workaround?

[Mobeen Iqbal]

Hello Everyone.

For some time, I have been using the AI content describer add-on version
2025.06.17 with NVDA 2025.3.2 with the default Pollinations (OpenAI)
model to describe images that I copy to the clipboard. However, recently
it has stopped providing image descriptions. The message I receive is:

"I’m not able to view images uploaded in this chat. If you can describe
it to me or share the key details or any text it contains, I’d be happy
to help interpret or discuss it!"

It doesn't seem to describe on-screen content consistently either. I
have tried google gemini and it's variations using my API key, but the
add-on just gives me a 404 error when I try to generate descriptions, no
matter what variation  of gemini I try.

I have also tried downloading ollama so I can run AI models, but image
descriptions don't seem to work with the local models either. I don't
mind using locally installed models as long as they can generate
detailed descriptions of pictures, such as extracting text or describing
people/objects in detail.

Before I file a bug report on Git Hub, Does anyone know of a workaround
or an alternative AI model I could use preferably low cost or free,
either locally installed or web based, or any alternative add-ons/apps
that I could use to describe  images?

Very best wishes,

Mo.

[muhammed ali çiçek]

Hello,
I have been experiencing the same issue for about two weeks now. I've tried it on both of my computers and I either get a similar response or no response at all.
So, I think the add-on is completely unusable. I suggest you try the CloudVision add-on as well. It can be found in the NVDA Add-on Store.
Best regards.
 

[Mobeen Iqbal]

Hi.

Thanks for the suggestion. I've just installed cloud vision, but I'm experiencing 2 issues. The responses seem to be in a different language, and they're not displayed on screen so you can read them back which is rather annoying. I've tried looking in the add-on settings to see if the language is set to English and it seems to be correct. I'll keep experimenting, it's such a shame that AI content describer hasn't been updated in some time. I'll wait a while to see if anyone else has any suggestions then I'll file a bug report.

Very best wishes,

Mo.

--
***
Please note: the NVDA project has a Citizen and Contributor Code of Conduct.
NV Access expects that all community members will read and abide by the rules set out in this document while participating in this group.
https://github.com/nvaccess/nvda/blob/master/CODE_OF_CONDUCT.md
 
You can contact the group owners and moderators via nvda-user...@nvaccess.org.
---
You received this message because you are subscribed to the Google Groups "NVDA Screen Reader Discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to nvda-users+...@nvaccess.org.
To view this discussion visit https://groups.google.com/a/nvaccess.org/d/msgid/nvda-users/CAMcewD%2Bxsx%3DrGzF5CvSojrOYMrS9DvXJX6rYxrD2RSRq0s2qFA%40mail.gmail.com.

[Gene Asner]

The description available in the Microsoft store says that if you press
NVDA control I twice quickly, a virtual review window will open.

Gene

>> <https://groups.google.com/a/nvaccess.org/d/msgid/nvda-users/CAMcewD%2Bxsx%3DrGzF5CvSojrOYMrS9DvXJX6rYxrD2RSRq0s2qFA%40mail.gmail.com?utm_medium=email&utm_source=footer>.

>
> --
> ***
> Please note: the NVDA project has a Citizen and Contributor Code of Conduct.
> NV Access expects that all community members will read and abide by the
> rules set out in this document while participating in this group.
> https://github.com/nvaccess/nvda/blob/master/CODE_OF_CONDUCT.md
> <https://github.com/nvaccess/nvda/blob/master/CODE_OF_CONDUCT.md>
>
> You can contact the group owners and moderators via
> nvda-user...@nvaccess.org.
> ---
> You received this message because you are subscribed to the Google
> Groups "NVDA Screen Reader Discussion" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to nvda-users+...@nvaccess.org

> <mailto:nvda-users+...@nvaccess.org>.

> To view this discussion visit

> https://groups.google.com/a/nvaccess.org/d/msgid/nvda-users/ecb4902a-2cd4-da8d-4993-f987c19c8043%40gmail.com
> <https://groups.google.com/a/nvaccess.org/d/msgid/nvda-users/ecb4902a-2cd4-da8d-4993-f987c19c8043%40gmail.com?utm_medium=email&utm_source=footer>.

[Andrew Downie]

Yes, pressing NVDA+ctrl+i twice quickly provides a virtual viewer.  The info is right there in the Cloud Vision description on the Addon Store.

I wrote to Carter about the problem with AI Content Describer but have not yet had a repsonse.

Andrew

To view this discussion visit https://groups.google.com/a/nvaccess.org/d/msgid/nvda-users/882699d5-315f-120b-db1d-a83f9f65528e%40gmail.com.