Deadline of 600.0s exceeded

Gemini API
1

Hi everyone. I’ve been having this problem with the Gemini API for a while now. I’m building an app that transcribes and summarizes recordings. The problem is that during transcription gemini takes more than 600 seconds to respond causing it to time out. Initially I thought it was because the recordings were 60 minutes or more long, so I started breaking the recordings into 15-minute chunks. I don’t think Gemini really needs 10 minutes to process a 15 minute transcript because when everything was working properly, The time to process a recording varied between 20 and 60 seconds.
Here is the full error:
Deadline of 600.0s exceeded while calling target function, last exception: 503 upstream request timeout

I want to point out that the recordings are valid and in MP3 format and that the problem occurs with both gemini 2.5 pro and gemini 2.5 flash.
Does anyone know how to help me?
Thank you all!

3

Hi @Ajxjencj_Cjdbcurv ,

Welcome to the AI Forum!!
This seems to be a temporary issue, could you please let us know if this issue still persists.

5

I’ve been struggling with this for weeks now. it’s almost impossible to use, even with catching errors and retrying. In my case, it’s just short audio segments: 30s / 200-500 characters. sometimes it’s

:cross_mark: TTS generation timed out for intro: Operation timed out after 180 seconds

sometimes its

  • ERROR - [_generate_with_timeout:826] - :cross_mark: Error in TTS generation: 500 INTERNAL. {‘error’: {‘code’: 500, ‘message’: 'An internal error has occurred. Please retry or report in

I have no choice but to switch back to openAI unless this is resolved

6

It will usually work eventually after 2-4 attempts on same generation, but it makes it very slow and much more expensive

7

seems to be a problem with both gemini-2.5-pro-preview-tts and gemini-2.5-flash-preview-tts

9

@Kinovate_Inc

Welcome to the AI Forum!!
This seems to be a temporary issue, could you please let us know if this issue still persists.