geminireset.

QUOTA FIELD NOTES · VERIFIED

Gemini API quota reset: daily limits and 429 errors

A Gemini API rate-limit error is not the same as a Gemini app or Antigravity allowance running out. Start with the limit named in the API response and your project in AI Studio.

Identify the exhausted limit first

Google documents separate limits for requests per minute, input tokens per minute and requests per day. Going over any applicable metric can stop a request even when other metrics have room. Model and project tier matter; inspect your active limits in AI Studio.

The API also has spend-based rate limits for some projects. A 429 response is not enough information to assume that waiting until tomorrow is the right fix. Read the error and identify which budget or rate window was exhausted.

Daily request quota resets at midnight Pacific

Google states that requests-per-day quota refreshes at midnight Pacific time. Pacific time changes with daylight saving, so do not hard-code a single UTC offset for the whole year. Minute limits and daily limits are different windows.

If you save a daily reminder here, convert the relevant midnight in America/Los_Angeles to your local date and time, then choose Gemini API. This is a manually entered reminder rather than an API-aware quota monitor.

A different API key is not a different quota

Rate limits apply to the project, not to each API key. Swapping keys within that project does not create more allowance. Check that you are inspecting the same project and model as the failing request.

Keep the error’s metric, model and project together when investigating a limit. Once you know the relevant window, set one reminder instead of repeatedly retrying against a still-exhausted budget. Antigravity reset announcements do not establish a change to your API project’s quota.

Original sources

Limits may change. Check your product and the official sources above. This is an independent hika.fyi community project, unaffiliated with Google.