On this page

Identify the quota first

Google documents request-per-minute and request-per-day limits for Gemini CLI, with different allowances by authentication method. A prompt may also trigger multiple model requests, so user prompts and provider requests are not always one-to-one.

Before comparing numbers with another developer, confirm whether both of you use the same login method and plan. Otherwise the comparison is likely meaningless.

Use /stats for the session

Gemini CLI's `/stats` command summarizes model usage, and the CLI can present a summary when the session ends. That is useful for tokens and requests inside the active session.

Daily quota is a different question. Pair session stats with the account's documented daily allowance and reset behavior. If you use an API key, Google AI Studio also shows project-level rate limits and usage.

Keep your usage visible while you work. Get Super Notchy for $39 ↗

Track the number you can act on

For interactive work, watch remaining daily capacity and the current session's request pace. For pay-as-you-go use, add token cost and project spending. Super Notchy can read Gemini token counts from local CLI activity and keep that signal beside other assistants.

When the pace is high, move simple transformations to a faster model, shorten repeated context, or pause parallel work. The reset time tells you whether routing or waiting is the better choice.

Sources and further reading