Direct answer
There isn’t one Gemini Developer API free-tier rate-limit number that applies to every model and project. Google says limits vary by model and are applied per project, not per API key. To see the limits relevant to your workload, use the “View your active rate limits in AI Studio” link in Google’s rate-limits documentation, and check the project your application uses. Don’t budget from a quota reported for a different model or account.
For production budgeting, treat free access as limited evaluation capacity, not guaranteed capacity. Google describes the free tier as intended for developers and small projects, and the paid tier as an option for production applications needing higher rate limits. That is not a promise of uninterrupted service or a specific quota. (Gemini API pricing)
How Gemini API free-tier limits work
This article concerns the Gemini Developer API, not usage limits in the consumer Gemini app or a Google AI subscription.
Google says rate limits are usually measured in three dimensions:
- RPM: requests per minute
- TPM: input tokens per minute
- RPD: requests per day
Exceeding an applicable limit can trigger a rate-limit error. Limits vary by model; additional limits can apply to particular models or tools. Google’s rate-limit documentation also describes image-per-minute and token-per-day limits for some use cases. There is no single quota number to apply to all models and projects. (Rate-limits documentation)
Limits are applied per project, not per API key. Google says RPD quotas reset at midnight Pacific time; don’t assume that timing applies to every other limit. (Rate-limits documentation)
To check the quota you should plan around:
- Identify the exact Gemini API model ID your application will call.
- Open Google’s rate-limits documentation and follow “View your active rate limits in AI Studio.”
- Confirm you’re checking the project your application uses, then review each displayed limit and unit—such as RPM, input TPM, or RPD.
- Check the project’s billing tier separately; quota and billing tier are different things.
A dated report for another model or project is not a substitute for checking your own active limits.
What happens when you hit a limit—and can you be charged?
Requests: Exceeding an applicable limit can trigger a rate-limit error. Google’s documentation identifies 429 RESOURCE_EXHAUSTED for spend-based rate limits; it recommends waiting and retrying after a short period for that case. This does not establish one reset interval for every quota or how every failed request is accounted for. (Rate-limits documentation)
If the project remains on the Free Tier: A Google AI Developers Forum responder wrote on November 18, 2025 that exceeding the free-tier quota without an enabled billing account would not generate a bill, and that calls would likely fail with 429 RESOURCE_EXHAUSTED. That is an attributed forum answer, not a complete billing-policy guarantee for every configuration. (Forum discussion)
If you upgrade to the Paid Tier: Google says upgrading means linking a billing account and prepaying at least US$5 (or equivalent) in credits. Its pricing page describes free input and output tokens under the Free tier and paid pricing for production applications. The official pages do not explicitly say whether a model-specific free allowance remains after Paid Tier activation. For a conservative budget, estimate usage at the selected model’s Paid Tier price and do not subtract a free allowance unless Google’s pricing page confirms it applies to your project and model. (Billing guide; pricing)
Free tier versus a production API budget
Google positions free access for developers and small projects getting started, while its paid tier is positioned for production applications that need higher rate limits and additional features. These descriptions support using the free tier for evaluation, but they do not promise that a particular paid tier will meet your application’s throughput or availability needs. (Gemini API pricing)
A practical decision rule:
- Prototype or internal test: Free access may be suitable if the model is available to your project and the workflow can tolerate quota limits or interruptions.
- User-facing production service: Check the project’s active limits against expected peak demand, then budget using the paid price for the exact model and features you intend to use. “Higher rate limits” is not a guaranteed capacity commitment.
- Workload that cannot tolerate throttling: Confirm the applicable production options and commitments for your use case rather than assuming that enabling billing alone guarantees capacity.
Example paid API cost estimate
Google lists Gemini 3.8 Flash at an introductory standard price through December 31, 2026 of US$0.75 per 1 million input tokens and US$3.75 per 1 million output tokens. Google says standard pricing becomes US$1.50 per 1 million input tokens and US$7.50 per 1 million output tokens on January 1, 2027. These are model-specific paid prices, not prices for every Gemini model or a guarantee that this model is included in your project’s free tier. (Gemini 3.8 Flash pricing announcement)
For a hypothetical workload of 1 million input tokens and 250,000 output tokens in a month, the introductory standard estimate is:
- Input: 1 × US$0.75 = US$0.75
- Output: 0.25 × US$3.75 = US$0.9375
- Estimated total: US$1.6875, or about US$1.69
This is a token-price calculation, not a measured bill. It assumes standard paid use at the stated introductory rates and excludes separately priced tools or other usage. At the standard rates scheduled from January 1, 2027, the same token volumes calculate to US$3.75. Recheck official pricing before making a longer-term forecast because prices and terms can change. (Gemini 3.8 Flash pricing announcement)
Bottom line
For the Gemini API free tier, check the active limits for your model and project in AI Studio; there is no universal request quota to budget against. Free usage can help with evaluation, but for production, compare expected demand with the project’s active limits and the current paid rate for the model you plan to use. Check the project’s billing tier separately: official guidance treats Free and Paid as separate tiers, but does not explicitly establish whether a free allowance remains after Paid Tier activation. Don’t assume that billing status alone guarantees production capacity or a free allowance. (Rate-limits documentation; billing guide; pricing)
Sources
Rate limits - Gemini API | Google AI Studio
Billing - Gemini API | Google AI for Developers
Gemini Developer API pricing | Google AI for Developers
What's new in Gemini 3.8 Flash | Google AI Studio
Quick question on Google AI Studio free tier limits (Gemini API) | Google AI Developers Forum