There isn’t one Gemini API free-tier limit that applies to every model and project. Google directs developers to check active limits for their project and model in Google AI Studio. Don’t budget around an old RPM, TPM, or RPD figure. (Google’s rate-limit documentation)
What are the Gemini API free-tier limits?
The Free tier gives eligible projects access to some models, up to each model’s free-tier rate limits. The relevant limits can include:
- RPM: requests per minute
- Input TPM: input tokens per minute
- RPD: requests per day
- Other model-specific limits, where applicable
Any applicable limit can stop requests independently. Google says RPD resets at midnight Pacific time and experimental and preview models have more restricted limits. (Google’s rate-limit documentation)
To find your numbers: open the AI Studio rate-limit view and check the project and model you actually use. Record each displayed limit; don’t assume another project or model has the same allowance. Google says limits are applied per project, not per API key. (Google’s rate-limit documentation)
Estimate whether your workload fits
Compare both request volume and input-token volume against the active limits. For example, if a project’s dashboard showed 30 RPM and 120,000 input TPM, and requests averaged 4,000 input tokens each, the token limit would allow about 30 requests per minute (120,000 ÷ 4,000), matching the RPM ceiling before accounting for daily limits or traffic variation. These are hypothetical inputs—not published Gemini free-tier quotas—and actual capacity may vary.
What changes when you enable billing?
Enabling billing moves the project from the Free tier to a Paid tier; don’t assume the project keeps its free-tier, free-of-charge token allowance. Google describes Paid-tier usage as priced by model and usage type. Check that model’s Paid tier pricing entry before linking billing. The Free tier’s limits are rate limits, not a free-token allowance that carries over to a billed project. (Google’s billing guide; Google’s pricing page)
For Google’s documented Free-to-Paid setup, upgrading means linking a billing account and adding at least $5 (or equivalent in another currency) in prepaid credits. Google’s billing guide also discusses Postpay and billing-plan transitions, so check which plan is available for your account rather than assuming the same payment setup applies to everyone. (Google’s billing guide)
To estimate token charges, use the selected model’s price for the request type you plan to use:
Estimated token cost = (input tokens ÷ 1,000,000 × input price per million) + (output tokens ÷ 1,000,000 × output price per million)
Use the model’s current pricing entry and confirm its billing units and conditions. Add any separately priced features or usage—such as caching or tools—where applicable. The estimate is not a complete bill if those charges apply. (Google’s pricing page)
Google’s billing guide lists these billing-tier caps and qualifications. They are billing-account controls, not RPM or token quotas for a project. Google says tiers and billing-account caps are determined at the billing-account level; its announcement describes the tier cap as a maximum monthly spend limit across that account. (Google’s billing guide; Google’s announcement on spend caps)
| Usage tier | Qualification shown by Google | Billing-account tier cap |
|---|---|---|
| Free | Active project or free trial; some projects need verification | N/A |
| Tier 1 | Set up and link an active billing account | $250 |
| Tier 2 | $100 paid and 3 days since first successful payment | $2,000 |
| Tier 3 | $1,000 paid and 30 days since first successful payment | $20,000–$100,000+ |
Google separately lists rolling 10-minute spend-based limits of N/A for Free, $10 for Tier 1, $50 for Tier 2, and $200 for Tier 3. Whether these apply depends on billing history and account standing; hitting one can return a 429 error even if request and token limits are not exceeded. Check the current billing and rate-limit pages for your account before using these figures in a budget. (Google’s rate-limit documentation)
Is the free tier suitable for production?
Free access can be useful for prototyping, but don’t treat it as guaranteed production capacity. Compare your expected peak request and input-token volumes with the active limits for your project and model, then decide whether the available capacity and billing controls meet your service needs.
Also check the data terms for your use case. Google’s pricing page describes Free-tier content as used to improve its products and Paid-tier content as not used for that purpose. But the Gemini API Additional Terms, effective March 23, 2026, state that for users in the European Economic Area, Switzerland, or the United Kingdom, the Paid Services data-use terms apply to all services—including unpaid Gemini API quota. Separately, those terms require Paid Services when making API clients available to users in those regions. The data-use condition and the rule about serving users are distinct; review the current terms for your location and deployment before sending sensitive data or serving users there. (Google’s pricing page; Gemini API Additional Terms)
Bottom line: the exact free-tier quota is the active limit shown for your project and model in AI Studio—not a single number to reuse across Gemini. Use those limits to size a prototype, and verify capacity, current model pricing, billing-plan options, spend controls, and applicable data terms before relying on the API in production.
Sources
Rate limits | Gemini API | Google AI for Developers
Billing | Gemini API | Google AI for Developers
Gemini Developer API pricing | Google AI Studio
Gemini API Additional Terms of Service
Giving you more transparency and control over your Gemini API costs