Google AI Studio
Build with the Gemini API, from prompt to production
About Google AI Studio
Google AI Studio is Google's web workspace for building with the Gemini API. It is where developers try Google's latest models, shape prompts, build and remix small apps, and then get an API key to take the result into production. The site sums it up as going from prompt to production with Gemini, Omni, Nano Banana and more. It lives on its own host, aistudio.google.com, separate from the consumer Gemini app, and it is now also the home of the Gemini API documentation.
You sign in with a Google account and land in AI Studio. From there you can open a model playground, start building an app, or browse the app gallery for a project to remix.
In the playground you pick a model, such as a Gemini text model, Nano Banana for images or Veo for video, and test prompts, settings and tools like Search grounding or function calling.
To build an app, you describe it in plain language.
- •Model Playground - Try Gemini 3, Nano Banana, Veo, Gemini TTS, Gemini Live, Omni and Lyria side by side, with tools such as Google Search and Maps grounding, code execution and function calling.
- •Vibe Coding and App Building - Describe an app and get a working version you can run in the browser and refine by prompting.
- •App Gallery to Remix - Featured projects, including multiplayer games, 2D and 3D games, interactive art and personal websites, can be remixed as a starting point for your own build.
- •API Keys and Usage - Create and manage Gemini API keys and monitor resource usage in one place before moving an app to production.
- •Official SDKs and Interactions API - SDKs for Python, JavaScript, Go, Java and C++, plus the Interactions API for stateful, multi-turn agents with managed tool execution and history.
Google AI Studio is aimed at developers who want to build on Google's models. Software engineers use it to compare models and prompts before writing integration code, then generate an API key and ship with the official SDKs. Startups use the free tier to prototype features such as image generation, transcription or voice agents without upfront cost.
Pricing
Usage based, no flat monthly plan
Free to use with a free tier of Gemini API tokens. Paid use is pay-as-you-go per token, priced by model; Enterprise through Google Cloud.
