Beginner

Getting Started with Google AI Studio

Access the platform, understand the interface, create your first prompt, and generate your Gemini API key in minutes.

Accessing Google AI Studio

Getting started is straightforward:

  1. Navigate to aistudio.google.com
  2. Sign in with your Google account (any Gmail or Google Workspace account works)
  3. Accept the terms of service if prompted
  4. You'll land on the main dashboard with options to create new prompts
💡
Good to know: Google AI Studio is free to use with generous rate limits. No credit card is required. You can start prototyping immediately after signing in.

Interface Overview

The Google AI Studio interface is organized into several key areas:

Left Panel — Navigation

  • Create New: Start a new freeform, structured, or chat prompt
  • My Library: Access your saved prompts and tuned models
  • Prompt Gallery: Browse pre-built prompt examples from Google

Center — Workspace

  • Prompt input area: Where you write your prompts and system instructions
  • Response area: Where the model's output appears
  • Examples table: (Structured prompts only) Define input/output pairs

Right Panel — Settings

  • Model selector: Choose between Gemini Pro, Flash, and other available models
  • Temperature: Controls randomness (0 = deterministic, 2 = very creative)
  • Top-K / Top-P: Fine-tune response diversity
  • Max output tokens: Limit response length
  • Safety settings: Configure content filtering thresholds

Creating Your First Freeform Prompt

Let's create a simple freeform prompt to see AI Studio in action:

  1. Click "Create new prompt" or the "+" button
  2. Select "Freeform prompt"
  3. In the prompt area, type a request like:
Write a brief explanation of how neural networks work,
suitable for a complete beginner. Use an analogy to
make it easy to understand.
  1. Click the "Run" button (or press Ctrl+Enter)
  2. Watch the model generate a response in real-time

Exploring Chat Mode

Chat prompts let you build multi-turn conversations:

  1. Create a new "Chat prompt"
  2. Add a system instruction at the top (e.g., "You are a helpful coding tutor")
  3. Type your first message in the user input
  4. After receiving a response, continue the conversation with follow-up messages
  5. The model maintains context across the entire conversation

Structured Prompts

Structured prompts use example tables to teach the model patterns:

  1. Create a new "Structured prompt"
  2. Define your input and output columns (e.g., "Product Review" and "Sentiment")
  3. Add a few examples in the table:
    • "Great product, love it!" → "Positive"
    • "Terrible quality, broke immediately" → "Negative"
    • "It's okay, nothing special" → "Neutral"
  4. Add a test input and run to see the model apply the pattern

Getting Your API Key

When you're ready to use Gemini in your own applications:

  1. Click "Get API key" in the left navigation panel
  2. Click "Create API key"
  3. Select an existing Google Cloud project or create a new one
  4. Copy the generated API key and store it securely
⚠
Important: Never share your API key publicly or commit it to version control. Use environment variables or secret management tools to store it. If your key is compromised, regenerate it immediately from the API keys page.

Running Your First API Call

With your API key, you can test a quick call using curl:

curl "https://generativelanguage.googleapis.com/v1beta/models/gemini-pro:generateContent?key=YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "contents": [{
      "parts": [{"text": "Explain AI in one sentence."}]
    }]
  }'

💡 Try It: Create All Three Prompt Types

Create one of each prompt type in Google AI Studio: a freeform prompt asking for creative content, a structured prompt for classification, and a chat prompt with a system instruction. Notice how each type guides you toward different use cases.

Understanding when to use each prompt type is key to getting the best results from AI Studio!