Google AI Studio: What It Is, What’s Free, and How It Differs From the Gemini App

Google AI Studio is Google’s free, browser-based workspace at aistudio.google.com where you test Gemini models, tune prompts, get a Gemini API key and build small apps. The free tier in Google AI Studio still lists Gemini 3.8 Flash as free of charge, while since 9 October 2026 the free Gemini app runs on Auto, which sends most prompts to Flash-Lite and gives you no say in the model.

That gap makes AI Studio look like a free upgrade to the Gemini app. The fine print says otherwise. Google’s own terms say Google AI Studio is for developers, “not for consumer use”. Human reviewers may read free-tier prompts. Image, video and Pro models need a paid key and a $5 prepaid minimum. This guide covers what Google AI Studio is, what is free, how to get a key, and where it differs from the Gemini app.

The Key Takeaways

• What it is: a free web tool for developers to test Gemini models, get an API key and build apps, launched on 13 December 2023.
• What is free: Gemini 3.8 Flash and 3.6 Flash, plus the Flash-Lite models. Nano Banana images, Veo 3.1 video, Gemini 3.1 Pro Preview and Gemini 4 Argon are not.
• Privacy catch: on the free tier, Google can use your prompts to improve its products and human reviewers may read them. The EEA, Switzerland and the UK are the exception.
• Billing catch: upgrading means prepaying at least $5. Credits expire after 12 months, and the $300 Google Cloud trial credit no longer covers AI Studio.
• Who it is for: Google’s terms say 18+, for professional or business use. The Gemini app remains the consumer product.

What Is Google AI Studio?

Google AI Studio is a free, web-based developer tool for trying Gemini models and then using them in your own code. Google introduced it on 13 December 2023, the same day the Gemini API went public. Google described it as a tool that lets you “quickly develop prompts and then get an API key to use in your app development.” It runs in any browser at aistudio.google.com and needs only a Google account.

Google AI Studio does three jobs. Each one has its own part of the interface.

The Playground: test prompts against any Gemini model

The Playground is the screen that opens by default. It looks like a chat window, but it gives you controls the Gemini app hides. You choose the model at the top. The Run settings panel adjusts model parameters and safety settings and holds the system instructions that shape every reply. It also switches on tools such as structured output, function calling, code execution and grounding. When a prompt works, the Get code button turns it into code in your preferred language.

Projects and API keys

Every Gemini API key in Google AI Studio belongs to a Google Cloud project, and the project carries the billing, usage limits and permissions. New users get a default project and API key created automatically after they accept the terms. The dashboard also shows usage, rate limits and spend for each project.

Build mode

Build mode turns a plain-English description into a working app. Google calls it “vibe coding”. It produces a full-stack web app or a native Android app, with a live preview beside the chat. Publishing a Build mode app runs on Google Cloud, with its own free allowance and limits.

Is Google AI Studio Free?

Yes, Google AI Studio is free to use with a Google account and no credit card, but only for the models Google puts on its free tier. Google’s billing FAQ is explicit: AI Studio usage “remains free of charge unless users link a paid API key for access to paid features.” Once you add a paid key, AI Studio usage on that key is billed.

Which models are free in Google AI Studio

Google’s Gemini API pricing page marks each model as either “Free of charge” or “Not available” on the free tier. As of 23 October 2026, the split looks like this.

Two details catch people out. First, Google AI Studio is no longer a free route to Nano Banana: every image model needs a paid key, as our guide to whether Nano Banana is free explains. Second, since 18 September 2026 Google has limited the older Gemini 2.5 models to people who already used them. A new account should start on 3.8 Flash or 3.5 Flash-Lite.

Where to find your free tier limits

Google no longer prints free-tier request limits in its documentation. The rate-limits page sends you to the Rate limit page inside AI Studio, which shows the live numbers for your own project. Limits apply per project, not per key, and the daily count resets at midnight Pacific time. Any blog that quotes an exact daily figure is quoting a number Google does not publish, so check your own dashboard instead.

For the full price list and how the API compares with the Google AI plans, see our Gemini pricing guide.

What Google Does With Your Prompts in Google AI Studio

On the free tier, Google uses what you type into Google AI Studio to improve its products, and human reviewers may read it. The Gemini API terms say this directly. When you use free services “including, for example, Google AI Studio and the unpaid quota on Gemini API”, Google uses your content and the responses “to provide, improve, and develop Google products and services and machine learning technologies.”

The same section adds that “human reviewers may read, annotate, and process your API input and output.” Google says it disconnects the data from your account, API key and project before reviewers see it. Its advice is blunt: “Do not submit sensitive, confidential, or personal information to the Unpaid Services.”

Three situations change the rules:

1. You are in the EEA, Switzerland or the UK. The paid-service data terms apply to Google AI Studio and the free API quota there, even though you pay nothing.
2. Your account has a Cloud project with active billing, or is a Workspace enterprise account. Google AI Studio then counts as a paid service for that account, even when the specific prompt costs nothing.
3. You use a paid API key. Google does not use the prompts or responses to improve its products. It logs them for a limited period to detect abuse.

The practical rule: treat free Google AI Studio like a public test bench. Paste a sample contract, not your client’s real one.

Google AI Studio vs the Gemini App

Google AI Studio is a developer console; the Gemini app is Google’s consumer chatbot. Google’s terms draw the line themselves: use of Google AI Studio and the Gemini API “is for developers building with Google AI models for professional or business purposes, not for consumer use.” The table draws on Google’s own help pages for both products.

The free Gemini app has changed more than AI Studio this autumn. Google’s help page confirms that since 9 October 2026 people without a plan get Auto: most prompts go to Flash-Lite, prompts that need deeper reasoning may go to Flash or Pro, and there is no way to pick the model yourself. AI Plus keeps Flash but loses Pro, on a date Google sends each subscriber by email. Pro stays with AI Pro and Ultra, which since 13 October also get Gemini 4 Argon. Our explainer on whether Gemini is free covers the app side in detail, and Gemini’s usage limits explains the five-hour windows.

Which one should you use?

Use Google AI Studio if you write code, want an API key, or need to compare exact model versions with your own system prompt. Use the Gemini app if you want a chatbot that runs on your phone and needs no setup. Using AI Studio as a free daily chatbot works on a technical level. It also runs against Google’s own “not for consumer use” terms, and every free prompt is open to human review.

If you just want Gemini on a Mac

Many people open AI Studio because they want a better Gemini model than the free app gives them, not because they plan to write code. For that job, a key and a prepaid balance are the wrong tools. Fello AI puts Gemini in a native Mac, iPhone and iPad app alongside GPT, Claude, Grok, Perplexity and DeepSeek. It costs $11.99 a month, with a free tier to try it. There is no API key to secure and no credit balance to watch.

How to Get a Gemini API Key in Google AI Studio

Getting a Gemini API key in Google AI Studio needs a Google account and no credit card. Follow these steps from Google’s API key guide:

1. Go to aistudio.google.com and sign in with your Google account. You must be 18 or older and in a supported country.
2. Accept the Google Terms of Service and the Generative AI Additional Terms of Service.
3. New accounts get a default Google Cloud project and an API key automatically. Open Dashboard > API keys to see it.
4. To make another key, click Create API key and choose the project it belongs to.
5. Copy the key and store it as the environment variable GEMINI_API_KEY. On a Mac with the default zsh shell, add export GEMINI_API_KEY=your_key to ~/.zshrc.
6. Test it with the Get code button in the Playground, which turns a prompt you have already run into code.

What changed with API keys in 2026

Google tightened key security three times in 2026. From 7 May, unrestricted keys left unused for a long period are blocked. From 28 May, every new key created in AI Studio is an authorization key bound to a service account. From 19 June, the Gemini API rejects requests from unrestricted standard keys, while standard keys with explicit restrictions keep working. An older key marked “Unrestricted” needs the Restrict to Gemini API only option on the API keys page, or a replacement.

Treat the key like a password. Google’s guide warns never to commit it to Git or ship it inside a web or mobile app, where anyone can extract it. Whether a pay-per-use key beats a flat subscription for your own use is a separate question, answered in our breakdown of an API key versus a $20 subscription.

How Google AI Studio Billing Works When You Upgrade

Upgrading Google AI Studio to the paid tier means linking a Cloud Billing account and prepaying at least $5 in credits. Google introduced Prepay and Postpay plans on 23 March 2026, and new users now default to Prepay. The rules are stricter than most people expect.

The prepaid credit rules

You buy credits on the AI Studio Billing page. The minimum purchase is $5 and the balance tops out at $5,000. Credits expire 12 months after purchase and are not refundable, except when you switch the account to Postpay. When the balance hits $0, every API key on that billing account stops at once, and requests fail with an HTTP 402 Payment Required error. Your projects do not fall back to the free tier on their own; you have to top up, or disable billing on each project to return it to free.

The $300 Google Cloud welcome credit no longer helps here. Google’s billing page says that accounts opened after 2 March 2026 cannot spend the welcome credit on the Gemini API or AI Studio. Credit granted before that cutoff can still be spent until it expires. Google’s free-trial page puts it plainly: “The $300 credit can't pay for Gemini API in AI Studio costs.” The credit still works on Gemini Enterprise Agent Platform, the Google Cloud product formerly called Vertex AI. It serves the same Gemini models with standard Cloud billing.

Usage tiers and spend caps

You can also set a lower monthly cap per project on the Spend page. Google labels project caps experimental and warns of overages during a billing delay of about 10 minutes, so long batch jobs can run past the cap.

Build Mode: Making Apps in Google AI Studio

Build mode in Google AI Studio creates a complete app from a text prompt. For web apps it generates a React front end and a Node.js server. For Android it generates a Kotlin and Jetpack Compose project you can run in a browser emulator. Google says Build mode runs on the Antigravity Agent, the agent inside Google Antigravity, which manages files across the project and verifies its own code changes.

You keep refining the app in the chat panel, or by pointing at part of the preview and describing the change. Code can sync both ways with a GitHub repository or download as a ZIP. Build mode stores your Gemini API key on the server side, so it never ships to visitors’ browsers.

Publishing is where costs start. Eligible accounts get the Google Cloud Starter Tier, which publishes up to two apps to Cloud Run with no billing account and an optional your-app.ai.studio address. Google Workspace accounts and anyone with a current or past paid Cloud billing account are not eligible. Shared and published apps also spend your own quota: Google notes that when you share an app, “API calls count toward your usage limits.”

Google AI Studio Not Working? Status Page and 403 Errors

When Google AI Studio fails, check the official AI Studio status page first. Google logs incidents there for both AI Studio and the Gemini API. On 1 October 2026, for example, it reported voice-cloning requests failing with a 404 error and marked the issue resolved the same afternoon.

A 403 Access Restricted error is different. Google’s troubleshooting page says it means you are using AI Studio in a way the terms do not allow, most often from an unsupported region. Two other causes are age-related: you are under 18, or your Google account has not verified your age. In Google Colab, the check uses the location of the Colab server, not yours.

A No Content reply means the response was blocked. Hover over the warning to see whether a safety setting caused it, and loosen the setting in Run settings if your use case allows.

The Verdict

Google AI Studio is the best free way to test Gemini models, and since 9 October it lets you choose a stronger model than the free Gemini app guarantees. That makes it the right first stop for anyone who writes code or needs an API key. It is the wrong place for private work on the free tier, and it is not built to be your everyday chatbot. Use it to prototype, keep sensitive material out until you pay, and set a project spend cap before you buy your first $5 of credit.

FAQ