Gemini AI Studio: What is free, what costs money and how it differs from the Gemini app

Plans, prices and features checked on the official pages on October 24, 2026.

Gemini AI Studio is Google AI Studio, a web workspace at aistudio.google.com where developers and curious users test Gemini models, write prompts, build small apps and create an API key. It is a different product from the Gemini app, which is the consumer assistant. AI Studio is free to start, but Google’s pricing page says content sent on the free tier may be used to improve Google’s products, and the newest and most capable models have no free tier.

Everything below comes from Google’s own documentation, read on October 24, 2026.

What Google AI Studio is and who it is for

Google’s developer documentation describes AI Studio as a place to “test prompts, manage your API keys, monitor usage, and build prototypes.” It is the front door to the Gemini API, the service that lets software call Gemini models for text, images, audio, video and more.

The audience is mostly developers, students, researchers and product teams who want to try a model before writing code. Nobody has to code to open it, though. A Google account is enough to start typing prompts.

The Gemini API itself is documented as supporting image and video generation, long-context analysis, structured JSON output, function calling, built-in tools such as search, URL context, maps, code execution and computer use, real-time voice through the Live API, speech-to-text, text-to-speech and embeddings.

Key features: Prompting, Build mode and API key management

AI Studio bundles several jobs in one place:

• Prompt testing. Pick a model, write a prompt, change settings and compare results before you commit to code.
• API keys and projects. The API keys page creates and lists keys, and the Projects page shows your billing tier.
• Usage and limits. The rate limit page at aistudio.google.com/rate-limit shows your active limits.
• Build mode. You describe an app in plain language and get a live preview.

According to Google’s Build mode documentation, web apps default to a React frontend with a Node.js server runtime, and Android apps use Kotlin and Jetpack Compose. The Antigravity agent keeps context across prompts and manages multiple files. Apps can use npm packages, Firebase Firestore and Authentication, and Google Workspace integrations such as Gmail, Sheets and Drive.

You can start from a prompt, use “I’m Feeling Lucky” for an idea, remix an app from the App Gallery or import a GitHub repository. Deployment options are Cloud Run, two-way GitHub sync or a ZIP download. For web apps, Google says a GEMINI_API_KEY is set up automatically and kept server-side, so it is not exposed in client code. For more on the coding agent behind Build mode, see our report on Claude Opus 5.5 in Google Antigravity.

How AI Studio differs from the Gemini app

The two products share the Gemini model family but serve different people.

The billing models are separate. A Google AI Pro subscription pays for the Gemini app, while API usage is billed through a Cloud Billing account. Treat the two as independent unless Google states otherwise on the page you are looking at.

If you mainly want to chat, edit photos or make slides, the app is the better fit. Our guides to Gemini photo prompts and Gemini Notebook cover those app features.

What is free in the Gemini API and what is not

Google’s billing documentation says new accounts start on the free tier, which gives access to certain models under free-tier rate limits. The pricing page lists free input and output for many models, including Gemini 3.8 Flash, Gemini 3.6 Flash, Gemini 3.5 Flash-Lite, Gemini 3.1 Flash-Lite, Gemini 2.5 Pro and Gemini 2.5 Flash. Gemma 4 and Gemini Embedding 2 are also free to use.

Several models have no free tier at all. These include Gemini 4 Argon, Gemini 3.1 Pro Preview, Gemini Omni Flash, the Nano Banana image models and the Lyria music models. Google also says that Google Cloud welcome credits cannot be used for Gemini API or AI Studio usage.

The catch is data use. On the free tier, the pricing page lists content as used to improve Google’s products for nearly every model. Google’s billing page says paid use keeps prompts and responses out of product improvement, and that AI Studio prompts fall under the same paid-services terms if at least one of your API projects has billing enabled. Check the Terms of Service before pasting anything private or confidential into a free-tier project.

Plans and pricing: What the paid API costs

The API has no flat subscription. Paid use is billed per million tokens, and Google’s billing page says cost depends on input tokens, output tokens, cached tokens and how long cached tokens are stored. The table shows a selection of Standard-tier prices from Google’s pricing page.

Prices checked on October 24, 2026. Batch, Flex and Priority tiers are priced separately, and image, audio and video models use their own units. Our report on Gemini 4 Argon for developers covers that model’s launch price.

Free-tier request limits are not listed in the pages used here. Google says they are shown per model on the rate limit page inside AI Studio, so look there for current numbers.

How to get started and get an API key

Google’s documentation gives this route:

1. Sign in to Google AI Studio with a Google account.
2. Open the API keys page. Google says AI Studio automatically creates a project and key for new users. To make another, click Create API key.
3. Set the key as an environment variable named GEMINI_API_KEY, for example export GEMINI_API_KEY="YOUR_API_KEY" on Linux or macOS. On Windows, add it as a user variable and open a new terminal.
4. Install an SDK: pip install -U google-genai for Python or npm install @google/genai for JavaScript.
5. Send a first request. Google’s quickstart uses the model gemini-3.8-flash and a plain REST endpoint at generativelanguage.googleapis.com with the key in an x-goog-api-key header.

Google lists these security rules for the key:

• Treat it like a password.
• Never commit it to Git.
• Do not embed it in client-side web or mobile apps; route calls through a backend instead.
• Use a secret manager in production.
• Set billing alerts.
• If it leaks, create a replacement, deploy it, then disable the old key.

If the Create API key button is unavailable on a work account, Google says a Cloud administrator may need to grant permission.

Limits, tiers and the move to paid use

Rate limits are measured in requests per minute, input tokens per minute and requests per day. Exceeding any one triggers an error. Google says limits apply per project, not per key, and daily quotas reset at midnight Pacific time.

To go paid, click Set up billing on the API keys or Projects page, then link a billing account. Google’s quickstart says this requires prepaying at least $5. The billing page says a prepay purchase can be $5 to $5,000, and that requests fail with HTTP 402 if the balance reaches $0. A postpay option charges at the end of the month or when you hit your spend cap.

Build mode has its own caveats. The documentation does not describe a free tier for it, says calls from shared apps count toward your usage limits, and notes that Cloud Run deployment pricing may apply. Apps are private by default, and there is no built-in storage or real-time collaborative editing.

Alternatives to AI Studio for different jobs

AI Studio is not the only way to reach Gemini or other models.

• The Gemini app. Pick it for chat rather than development. Google’s subscription page lists a Free plan at 0 euros, Google AI Plus at 4.99 euros per month, Google AI Pro at 21.99 euros and Google AI Ultra from 99.99 euros. Those prices were shown in euros on the page read, and US prices may differ.
• Google Cloud. Teams that need Google Cloud governance can reach partner models there. See our coverage of OpenAI models in Model Garden.
• Other providers’ consoles. Anthropic and OpenAI run their own developer platforms with separate keys and billing.

Common questions