The gemini-2.5-flash-preview-04-17 model works fine in Aistudio for normal conversations.
However, when tested via the API, a large number of 500 and 503 errors occurred.
Out of 15 data points, only 1 was successful. The remaining 14 failed after 3 retries each, resulting in a total of 42 failed attempts.
All error messages were 500/503.
Is the model currently under maintenance?
When testing with Gemini 2.0 Flash, everything works normally.
1 Like
Gemini-2.5-Flash-Preview-04-17-thinking also has the same problem.
same problem for me.
@sam_hong , @usaki_Huang , @fu_guo
Thank you for reaching out in the forum. Welcome to the community.
Gemini team were making a lot of small changes earlier this week in light of IO.
This issue should be fixed now, please let us know if these issues still persists.
Hi,
Is it fixed? Still getting frequent 5xx errors:
Error calling Gemini API: 500 Internal Server Error on POST request for “https://generativelanguage.googleapis.com/v1beta/models/gemini-2.5-flash-preview-04-17:generateContent”: “{ “error”: { “code”: 500, “message”: “An internal error has occurred. Please retry or report in Troubleshooting guide | Gemini API | Google AI for Developers”, “status”: “INTERNAL” }}”
Anyone knows if this is a problem with only free tier or with all tiers? I am using REST API directly from my springboot application. Also will vertexai sdk APIs fare better (may be because of billing)?