Quick answer
The Gemini API has a Free tier, but it covers only certain models and comes with model- and project-specific limits. Google documents those limits as requests per minute (RPM), input tokens per minute (TPM), and requests per day (RPD); the active values can be viewed in Google AI Studio. Limits are applied per project, not per API key, and daily request quotas reset at midnight Pacific time. (ai.google.dev)
The free tier can suit evaluation and small projects, but don’t treat it as a dependable production budget: Google describes its paid tier as intended for production applications needing higher volumes and features. Free-tier prompts and responses may be used to improve Google products; paid-tier prompts and responses are not. (aistudio.google.com)
For an illustrative paid-cost estimate, Gemini 3.8 Flash’s introductory standard rates are $0.75 per 1 million input tokens and $3.75 per 1 million output tokens through December 31, 2026. For example, 200 million input tokens plus 50 million output tokens would cost $337.50 before any eligible free usage, taxes, or other charges. (aistudio.google.com)
Information below reflects Google documentation available on November 8, 2026. Model availability, prices, and limits can change.
What does the Gemini API free tier include?
Google’s pricing page describes the Free tier as offering access to selected models, free input and output tokens for eligible usage, and Google AI Studio access. It does not mean every Gemini model or feature is free. For example, Google says Gemini 4 Argon is not available on the Free tier. Check the pricing entry for the specific model ID you plan to call; a model family name or an old quota figure does not establish current eligibility. (aistudio.google.com)
Google does not provide one universal quota that applies to every model and project. It measures limits in dimensions such as:
- RPM: requests per minute
- TPM: input tokens per minute
- RPD: requests per day
Limits vary by model and usage tier. Some models or features may use additional measures, such as images per minute or tokens per day. Exceeding any applicable limit can result in a rate-limit error. (ai.google.dev)
To see the limits that apply to your project: open Google AI Studio’s rate limits view and check the model you intend to use. Google applies these limits per project rather than per API key, so creating extra keys within the same project does not give you a separate project quota. (ai.google.dev)
Historical third-party reports about particular models or daily request counts should not be used as current limits: quotas differ by model and account, and Google directs developers to their active limits in AI Studio. Likewise, quotas for the consumer Gemini app or access through a Google AI subscription are not the same thing as Gemini API billing and limits. (ai.google.dev)
Is the free tier suitable for production?
Sometimes for a small, low-volume workload—but it is not the safest default for a production service. Google positions free access for developers and small projects getting started, while its paid tier is for production applications requiring higher volume and advanced features. The free tier’s quota is a ceiling, not a promise that a particular workload will have the capacity or continuity it needs. (aistudio.google.com)
There is also a data-use trade-off. Google’s pricing page says free-tier content may be used to improve its products, whereas content on paid services is not used for that purpose. Avoid sending sensitive or confidential information through free-tier requests unless your own review of the applicable terms and data-handling requirements supports doing so. (aistudio.google.com)
Google’s terms, effective March 23, 2026, say the Gemini API and AI Studio are for developers building for professional or business purposes, rather than consumer use. They also require Paid Services when making API clients available to users in the European Economic Area, Switzerland, or the United Kingdom. That regional requirement is separate from whether a project’s request volume fits a free quota. (ai.google.dev)
Practical choice: use the free tier for experiments or modest workloads where its current project limits and data terms are acceptable. Plan for paid usage when your service needs higher limits, uses data you do not want used to improve Google products, or must serve users in the EEA, Switzerland, or UK. Google’s documentation does not establish that merely enabling the API or having free quota guarantees production reliability.
How much might paid Gemini API usage cost?
Gemini API charges depend on model, token counts, and applicable pricing conditions—not on a consumer Gemini app subscription. Google’s Gemini 3.8 Flash announcement gives introductory standard rates of $0.75 per 1 million input tokens and $3.75 per 1 million output tokens through December 31, 2026. Recheck the official pricing page before committing to a budget, especially for usage after that date or for other models and modalities. (aistudio.google.com)
Worked example: 100,000 text requests
Assume each request averages 2,000 input tokens and produces 500 output tokens:
- Input: 100,000 × 2,000 = 200 million tokens
- Output: 100,000 × 500 = 50 million tokens
- Input cost: 200 × $0.75 = $150
- Output cost: 50 × $3.75 = $187.50
- Estimated total: $337.50
This is a standard paid-tier estimate for Gemini 3.8 Flash at its introductory rates. It assumes every listed token is billed at those rates and excludes free-tier usage, taxes, discounts, and other services or tools. Actual costs depend on your model, usage, and pricing conditions. Google’s pricing page also describes features such as Batch API and context caching, but whether they reduce your bill depends on your workload and eligibility. (aistudio.google.com)
Google’s billing documentation says upgrading to a Paid tier involves linking a billing account and prepaying at least $5 (or the equivalent in another currency). That is a minimum prepayment to access the paid tier, not the estimated price of the workload above. Billing options and account transitions can vary, so review the billing page for your project before upgrading. (ai.google.dev)
What to check before launching
- Confirm the model: verify the exact model ID is available on the Free tier or identify its paid rates in Google’s pricing documentation. (
aistudio.google.com) - Check your project’s active quota: review RPM, TPM, and RPD for that model in AI Studio; do not rely on old quota tables or figures from other projects. (
ai.google.dev) - Review data terms: decide whether the Free tier’s data-use terms are appropriate for the prompts and responses your application handles. (
aistudio.google.com) - Estimate paid usage: multiply expected input and output tokens by the applicable per-token rates, then account separately for features, modalities, or conditions that affect pricing. (
aistudio.google.com) - Check regional rules: if your API client will be available to users in the EEA, Switzerland, or UK, Google’s terms require Paid Services. (
ai.google.dev)
Sources
Gemini Developer API pricing — Google AI Studio