429 RESOURCE_EXHAUSTED on first use after a long idle period

What this teaches

How to 429 RESOURCE_EXHAUSTED on first use after a long idle period.

The problem

Issue googleapis/python-genai#17 (closed, 20 comments): I am encountering a ClientError: 429 RESOURCE_EXHAUSTED error when attempting to use the gemini-2.0-flash-exp model with the Google Search tool. This error occurs immediately upon sending the first message in a chat session, even after a period of approximately 48 hours of inactivity, indicating that this is not a result of exceeding typical usage quotas. Environment: Python, Windows 11, google-genai 0.2.2. Expected Behavior: The `gemini-2.0-flash

What worked (verified answer)

A 429 RESOURCE_EXHAUSTED on your first Gemini request after a long idle period is not your code hammering the API. On Vertex AI it means you hit a prediction-request quota, per a Google maintainer in the thread. Check your model quotas in the Cloud console and contact Vertex AI support. No client-side retry setting or SDK config fixes it, so do not burn time tuning retries.

Source

gh:googleapis/python-genai#17 - https://github.com/googleapis/python-genai/issues/17

In gh:googleapis/python-genai#17, the reporter ran into: Issue googleapis/python-genai#17 (closed, 20 comments): I am encountering a ClientError: 429 RESOURCE_EXHAUSTED error when attempting to use the gemini-2.0-flash-exp model with the Google Search tool. This error occurs immediately upon sending the first message in a chat session, even after a period of approximately 48 hours of inactivity, indicating that this is not a result of exceeding typical usage quotas. Environment: Python, Windows 11, google-genai 0.2.2. Expected Behavior: The `gemini-2.0-flash The verified answer was: A 429 RESOURCE_EXHAUSTED on your first Gemini request after a long idle period is not your code hammering the API. On Vertex AI it means you hit a prediction-request quota, per a Google maintainer in the thread. Check your model quotas in the Cloud console and contact Vertex AI support. No client-side retry setting or SDK config fixes it, so do not burn time tuning retries. Source issue: https://github.com/googleapis/python-genai/issues/17