Multiple models returning 503 on Gemini API

I am continuously getting 503 Service Unavailable errors with the Gemini API.

・ Environment: Google Apps Script / Free tier
・ Affected models: gemini-3.7-flash, gemini-3.8-flash, gemini-3.6-flash
・ Even after retrying and switching models, it still results in a 503.
・ The usage history records some output tokens, numerous 503s, and a few 404s.
・ In AI Studio, a “permission denied” error also occurred separately.

Please let me know what I should check, such as project settings, permissions, or model eligibility.

Hello @hiro2 ,

503 errors are due to our services being temporarily overloaded. Please see this pinned post for suggestions on how to handle this.