How to Check Your Gemini Usage

API active rate limits: per project, not per key

Google's Gemini API documentation states that active rate limits are viewed in AI Studio, and that limits are usually measured on three axes: requests per minute (RPM), input tokens per minute (TPM), and requests per day (RPD). Exceeding any one of the three triggers a rate-limit error.

The boundary that trips people up: limits apply per project, not per API key, and the daily request count resets at midnight Pacific time. Two keys in the same project share one quota. Google also notes that limits vary by model, usage tier, and account status, and that specified limits are not guaranteed, actual capacity can vary.

Documentedai.google.dev: Gemini API rate limits· verified 2026-08-20

API usage

For how much you've actually called the API, as opposed to how close you are to a limit, Google's billing documentation points to a different path in the same product: AI Studio's Dashboard > Usage view.

Documentedai.google.dev: Gemini API billing (usage view)· verified 2026-08-20

Cloud project quota and current usage

Every Gemini API key, including one created through AI Studio, belongs to a Google Cloud project, the per-project scope above applies to it either way. When you want that project's quota and current usage, rather than the active-limit view AI Studio shows for the key you're using, Google Cloud's quotas documentation is the authority: Cloud Console's IAM & Admin > Quotas & System Limits page lists quota usage and values for the selected project, and the dashboard for a specific API shows that API's quota information, including resource usage over time.

Documentedcloud.google.com: viewing and managing quotas· verified 2026-08-20Documentedai.google.dev: billing (every API key belongs to a Cloud project)· verified 2026-08-20

Costs

Gemini API costs can be viewed in Cloud Billing's cost-management pages. Google's billing docs say cost details are typically available within a day, but sometimes take more than 24 hours to show up, so a spend spike from the last few hours may not be visible yet even in the official dashboard.

Documentedai.google.dev: Gemini API billing (cost management, updated 2026-08-10)· verified 2026-08-20

Continuing Gemini CLI: a session snapshot, not a billing ledger

On an eligible route, Gemini CLI's own docs describe a built-in command: running /stats model gives a snapshot of the current session's token usage and information about the limits tied to your current quota, and a model-usage summary is presented when the session exits.

That boundary matters: /stats model and the exit summary describe this session, not your account's total usage or spend. For the account-level number, use the API usage, Cloud quota, or cost pages above instead.

Documentedgeminicli.com: quota and pricing docs (/stats model, updated 2026-06-18)· verified 2026-08-20

Where this guide stops

This page only routes you to the right official surface for a given usage question. It deliberately does not repeat the specific RPM/RPD/TPM numbers per model and tier. Those move, and Gemini rate limits explained owns that fact. It also doesn't compare tracking tools feature-by-feature. That comparison lives on best Gemini usage trackers for Mac. If you want a menu-bar view of your local Gemini CLI activity instead of checking these pages by hand, see SessionWatcher for Gemini.

Frequently Asked Questions

Where do I check my Gemini usage?

It depends which surface you mean: AI Studio for API active rate limits and usage, Cloud Console Quotas & System Limits for a project's quota, Cloud Billing cost management for spend, or /stats model for a Gemini CLI session snapshot on an eligible account.

Documentedai.google.dev: rate limits· verified 2026-08-20Documentedai.google.dev: billing· verified 2026-08-20Documentedcloud.google.com: quotas· verified 2026-08-20Documentedgeminicli.com: quota and pricing· verified 2026-08-20

Are Gemini API quotas per API key or per project?

Per project, not per API key, per Google's Gemini API rate-limits documentation. The daily request count resets at midnight Pacific time.

Documentedai.google.dev: rate limits· verified 2026-08-20

Is Gemini CLI still available for Google AI Pro, Ultra, or free Gemini Code Assist for individuals?

No, not since June 18, 2026. Continuing routes: Gemini CLI/IDE access through a Gemini Code Assist Standard or Enterprise license, Gemini Code Assist for GitHub through Google Cloud, and paid Gemini or Gemini Enterprise Agent Platform API keys.

DocumentedGoogle Developers Blog: Gemini CLI transition· verified 2026-08-20

Sources

Every claim above links its source at the point it is made. This table collects the same set with the date shown on the source page and the date it was checked for this guide.

SessionWatcher is an independent product and is not affiliated with, endorsed by, or sponsored by Google. Gemini is a trademark of Google LLC.