It depends on the product. Gemini Apps has standard access without a paid AI plan, subject to limits; that does not mean you can use the specific Gemini 2.5 Flash API model for free. As of November 9, 2026, Google says the API model is deprecated, unavailable to select for new use in AI Studio, and limited to existing users until December 1, 2026. Google’s documentation does not confirm here whether those existing users have a free Gemini 2.5 Flash allowance or what their current model-specific rate is. (aistudio.google.com)
Free access depends on where you use Gemini
| Product or access route | What is confirmed about free use |
|---|---|
| Gemini Apps | Google lists standard limits for use without an AI plan. Those app limits are not API access to Gemini 2.5 Flash. () |
| Gemini API / AI Studio | Google offers a free API tier with free input and output tokens for certain models. But the available documentation does not establish a free allowance for Gemini 2.5 Flash specifically. API access to this model is limited to existing users until December 1, 2026. (aistudio.google.com) |
| Paid or production API use | Google describes paid access as intended for production applications needing higher volumes and rate limits. The current model-specific Gemini 2.5 Flash rate is not confirmed in the pricing information available here. (aistudio.google.com) |
Bottom line on “free”: don’t assume that a general Gemini API free tier covers Gemini 2.5 Flash. The current free allowance for eligible existing users is unconfirmed here; new users should not assume they can select this API model. The Gemini Apps free-access statement applies to the app, not API calls.
What should you budget for the API?
Google’s June 17, 2025 pricing announcement gave Gemini 2.5 Flash rates of $0.30 per 1 million input tokens and $2.50 per 1 million output tokens. These are historical announced rates, not a verified current quote for existing API users. (developers.googleblog.com)
For scale, at those historical rates, 1 million input tokens plus 250,000 output tokens would cost $0.925: $0.30 for input and $0.625 for output. This is an arithmetic illustration based on the 2025 rates—not a current bill estimate.
Google says that beginning December 1, 2026, API requests using gemini-2.5-flash will automatically route to gemini-3.5-flash-lite and be billed at Gemini 3.5 Flash-Lite rates. The specific current input and output rates for that replacement are not confirmed in the pricing details available here, so don’t use the 2025 Gemini 2.5 Flash rates to forecast post-transition spend. (aistudio.google.com)
What changes on December 1, 2026?
Google’s model documentation says Gemini 2.5 Flash API access is limited to existing users until December 1, 2026. From that date, requests using its model ID are scheduled to route to Gemini 3.5 Flash-Lite and use that model’s rates. Google also lists migration changes, including function-calling and thought_signature validation, thinking_level replacing thinking_budget, and the removal of some sampling parameters and prefilled model turns. Review the migration documentation if your integration relies on those features. (aistudio.google.com)
A free API allowance, if your account and model qualify, is not a guarantee of production capacity or a fixed production cost. Google says API limits vary by model and tier and are applied per project; check the limits and billing that apply to your project before relying on access.
Bottom line: Gemini Apps access without a paid AI plan is different from Gemini 2.5 Flash API usage. Google documents restricted existing-user API access and a December 1, 2026 routing change, but the model-specific free allowance and current applicable rates are not confirmed here. Treat the June 2025 rates as historical and verify your project’s current terms before budgeting.
Sources
Gemini Apps limits and feature availability