Frequent 429 RESOURCE_EXHAUSTED errors with gemini-embedding-001 since mid-April

Google Cloud Build with AI Generative AI & Foundational Models
, ,
1

I’ve been using gemini-embedding-001 for a RAG-based system. Since mid-April, I’ve been encountering the following error frequently. I had never seen this error before.

Error embedding content (RESOURCE_EXHAUSTED): 429 RESOURCE_EXHAUSTED. 
{
  'error': {
    'code': 429,
    'message': 'Resource exhausted. Please try again later. Please refer to https://cloud.google.com/vertex-ai/generative-ai/docs/error-code-429 for more details.',
    'status': 'RESOURCE_EXHAUSTED'
  }
}

I checked the documentation at the provided URL and reviewed the suggested retry logic.

Does anyone know if there have been recent changes to this embedding model or if new limits have been introduced? Or why this might suddenly be happening?

2

Same issue here when using gpt-oss-120b via VertexAI.