Selected Gemini models failing to execute in Google AI Studio
When attempting to prompt or execute tasks in Google AI Studio, specific model endpoints are completely unresponsive.
• Not Working: gemini-3.6-flash, gemini-3-preview, and gemini-3.1-flash-lite
• Working Normally: gemini-3.8-flash and gemini-3.5-flash-lite
This issue persists across both playground execution and basic API calls. Please verify if this is an undocumented deprecation rollout or an active service outage affecting regional endpoints.
Hello @Ceeya_Vuya ,
There is no active service outage or deprecation rollout affecting gemini-3.6-flash or gemini-3.1-flash-lite—both models are active and supported (see the Gemini models documentation and Gemini deprecations). Please note that gemini-3-preview is not a valid model endpoint ID; if you are calling a Gemini 3 Preview model, please use gemini-3-flash-preview or gemini-3.1-pro-preview (gemini-3-pro-preview has been shut down).
If gemini-3.6-flash or gemini-3.1-flash-lite are still failing to execute, could you please share:
- What specific error message or HTTP status code (e.g., 404, 429, 503) appears when execution fails?
- Which country or region are you connecting from?
- A minimal code snippet or cURL command (with your API key redacted) showing the failed call, or a screenshot of the error in Google AI Studio.
Any specific reason why you are not using 3.8 flash?