Hi Gemini Team,
We’re observing a serious regression in the time it takes for files to transition to ACTIVE status.
Previously, this process would complete in under 3 minutes—often within a minute. However, it now consistently takes 15 to 25 minutes, typically over 18 minutes, which is a drastic increase.
This delay is significantly impacting our workflow. Could you please look into this and let us know what’s going on?
Looking forward to your help.
Best regards,
Wenjun Wang
We are also experiencing the same for about 2 days. Files that took 5-10s are now taking 2-5 minutes.
I wonder if it has anything to do with the roll-out of Gemini to Google Drive files.
Thanks for flagging! Can you share some details on the file type & size? I’ll raise this with the team as well
I don’t think it’s caused by file size though.
The file size is really small, like 3MB.
File type is also normal, which is mp4.
Noted - thank you. We’re investigating the issue
This is a huge issue for us rn
@Vishal any fix for this? The latency is quite alarming. Its creating a significant performance issue for us. Process that takes less than a minute is now well over 10 mins.
Hey folks, to help us with our investigation if you’re experiencing this issue, can you please DM me your project # (found here: https://aistudio.google.com/app/apikey)?
it happened to me too, the file size really doesn’t matter, below 1MB file will take forever to process until it shows “FAILED”, I tried in both Google AI Studio and my demo app
Hi Vishal, you can re-produce this issue by using Google AI Studio, upload a short video under 10s, if you can’t see it then something went wrong with the quota for individual API key or for project or whatever, I didn’t try to re-produce this issue on a brand new API key but if your side can not re-reproduce, I think I have no choice but try a brand new API key
@Vishal
I generated a new API key and tried and the same issue was reproduced.
I don’t know if it’s helpful, but when it’s the first “run” in a while, as it was with this new key, it seems fine, but subsequent calls are slow. In my workflow, I do both small videos and images. The time for the images seem fine. The videos however will stay as “PROCESSING” for a very long time.
Hi folks we are having a similar issue. We use the async API to upload and infer- await client.aio.files.upload() . Upload into Gemini server is very fast - but files take a while for videos to turn ACTIVE
Yesterday we’d get stuck in PROCESSING for 10+ minutes. Today it seems slightly better, but still very variable across different videos uploads
@Vishal - Feel free to open your DMs so we can share our project too!
I just read a news about the new release for the upgrade of Gemini 2.5 Pro, I used Google AI Studio and did notice that the name of the model automatically switched to “Preview 05-06”, which was great, but then I tried to upload a video, it was fast for the upload and it was fast too to say “FAILED” for the token extracting process. So the same issue is happening even after the upgrade for 2.5 Pro, which means the issue is still there 
latest update from my side: it works again now, I just tested by uploading a 10 seconds long video, around 2MB in size, and it took a short time to upload and to extract that file (turned into ACTIVE status). My location is Singapore, hope your side will see the same thing. Enjoy coding!
Hey folks, we pushed a number of small improvements to improve the processing for the File API. Please let me know if you’re still running into any issues / higher latency
We’re still facing the same issue. It still gets arbitarily stuck at PROCESSING for 10+ mins. This behaviour fluctuates by hours. Please suggest what do we do to handle this uncertainity better.
Thanks!
Yeah it’s still pretty unpredictable at times we are using gemini 2.0 flash
and for us the step to upload a file to gemini takes up a long time, although this happens unpredictably sometimes quite fast else very slow
