Model: gemini-2.0-flash
Tier-1
Hi
I am using the Gemini Batch Mode to send batches of 8K requests (text gen only, no img), strictly sequentially, to avoid hitting token queue limits.
After successfully processing 51 batches, I am now hitting the 429 Resource Exhausted error. I have deleted all files previously uploaded, in case there was an issue with that, and I am certain my batches are below the 10M token limit. I have no idea why the batches are now being refused.
Has anybody encountered this before?
Any insights are appreciated,
Many thanks
C.
@porBulerias ,
i understand that you are giving 51 batches, do you have an estimate of how many total requests will that make, as in total number of model inferences is that under the limit of RPD for you tier
also if you share the detailed 429 error message it might have further information on which rate limit the reqests are being rejected.
Thankyou
Thank you @Akhilesh_Kambhampati. The 51 batches were sent over 4 days (10 + 10 + 10 +21). Each containing 8K requests. It was on the last day (29/7) that after sending 21 batches (168K requests), it came back with 429. This continued all day yesterday, and continues today:
Batch creation failed on attempt 1: ClientError - 429 RESOURCE_EXHAUSTED. {‘error’: {‘code’: 429, ‘message’: ‘You exceeded your current quota, please check your plan and billing details. For more information on this error, head to: https://ai.google.dev/gemini-api/docs/rate-limits.’, ‘status’: ‘RESOURCE_EXHAUSTED’, ‘details’: [{‘@type’: ‘type.googleapis.com/google.rpc.Help’, ‘links’: [{‘description’: ‘Learn more about Gemini API quotas’, ‘url’: ‘Rate limits | Gemini API | Google AI for Developers’}]}]}}.