I’m experiencing a 429 RESOURCE_EXHAUSTED error when using Google Search Grounding with the Gemini API.
Environment
Model: gemini-3.5-flash
API: Gemini API generateContent
Runtime: Google Apps Script
Tool: googleSearch
Tier: Free / current project tier
Testing performed
A minimal Gemini request without Search Grounding succeeds:
Prompt: Reply with exactly: TEST_OK
Result: HTTP 200
Response: TEST_OK
The same minimal request with Google Search Grounding enabled fails:
tools: [{ googleSearch: {} }]
Result:
HTTP 429
RESOURCE_EXHAUSTED
"You exceeded your current quota, please check your plan and billing details."
The request contains only a very small prompt, so this is not a TPM-heavy request.
My AI Studio usage/rate-limit dashboard does not show corresponding usage that would explain the failure.
This appears to isolate the problem to Search Grounding rather than the Gemini API itself.
Could you please confirm whether Google Search Grounding is currently unavailable/restricted for gemini-3.5-flash on my tier, or whether this is an issue with the project’s grounding quota?
I can provide the project number, timestamps, full API response, and request details privately if needed.
I invite developers, researchers, and users working with Google AI Studio to conduct an independent verification of the platform’s data deletion lifecycle and backend persistence.
Steps to Reproduce the Deletion Verification:
Create a session: Open Google AI Studio and start a new prompt / multi-turn chat session.
Delete the prompt: Trigger the deletion command on the prompt within the AI Studio user interface.
Empty the Trash: Navigate to Google Drive, locate the associated .json file in the Trash folder, and select ‘Empty Trash’ (permanent deletion).
Run recovery: After a period of time (e.g., several days), execute the official Google Drive File Recovery tool on the account.
Check AI Studio: Re-open Google AI Studio and click on the restored .json file. Observe whether the complete conversational memory, prompt context, and token history are reloaded from the backend servers.
Technical Background:
The purpose of this test is to verify whether triggering deletion in Google AI Studio executes a cascading backend purge (DELETE/DROP) on persistent storage, or merely performs a client-side disconnection by moving the frontend metadata pointer (.json) to the Trash while retaining the underlying interaction data on persistent servers.
Verifying whether data remains recoverable following explicit permanent deletion commands is an essential aspect of evaluating data lifecycle transparency and statutory compliance under GDPR Article 17 (Right to Erasure).