I’m trying to work with uploaded image, not asking for model to generate image, only ask it to extract text from image, but it keeps giving me this error, regardless of models. Any idea on why? The same process worked totally fine few days ago.
4 Likes
Update: this issue has been resolved as of April 4th.
I hope Gemini team have better communication with the community about the status of product. It’s okay that things are not working, but we need to know when it’s broken.
5 Likes
what was the solution?
2 Likes
Can’t generate anything relates to image, even just for text extraction
1 Like
Hi @cubesat,
Many fixes has been rolled out in the last few weeks.. I hope you are able to generate images now.
Please repost or DM if you are still facing issues.
1 Like
No. Today I just tried to analyse a video getting the same error
2 Likes
Welcome to the Google AI Forum! 
Can you elaborate on your approach that you tried and share the error message.
1 Like
I have been experiencing the same in the last 24 hrs, mine is simply from either pasting or uploading a screenshot.
Opened a new chat thread, trouble shooted advice confirmed to open a new chat thread. Experiencing the same in the new chat thread this evening.
Due to issues losing chat history in 05/06, (I assumed it was due to tokens and that it was deprecating), but the current version is experiencing these errors much earlier now.
Have also restarted thinking it was cpu as my fan is tripping during these errors. and cleared cache.
Errors recorded: Internal Error, Failed to generate content, then Google Drive errors also recorded.
Hope it helps.
3 Likes
Same here lots of “Failed to generate content: permission denied. Please try again.” and could not connect to drive issues.
4 Likes
i am also getting this error message, hope they fix it soon.
2 Likes
Couple of things that can get us through are: -
- Trying drag and drop instead of pasting the image, regular workarounds like clearing the cache and cookies or the use of incognito mode.
- Another thing is if the image is large, resizing it to under 1MB file.
But if it’s urgent, we can move to Google Cloud Vision API for OCR or use VertexAI with OCR preprocessing as well.
4 Likes
scaling anything at all without proper multi-channel flow for notifications to end-user… i know that because its free we get duped to believe it makes us considerate to ignore this because fancy equipment and fancy big tech resources are being invested, but the multi-channel accountability to keep the user aware of changes or status is one of those structural obligations that make the “move fast, break things” approach very much an amateur product management policy
an AGILE expert from SCRUM told me that
Hi, I encoutered the same error. Is there any way to fix it?
I am getting the same error as on 28th Aug
1 Like
The bug came back today
Facing same issue today (5th Sept)
I am getting the same error as on 11th Sep '25
Getting the same error here.