Beginner

Getting Started with Gemini

Learn how to access and use Google Gemini through the web interface, Google AI Studio, and the API. Start your first conversation and explore multimodal inputs.

Ways to Access Gemini

Google provides several ways to interact with Gemini, each suited to different needs:

Access Method Best For Cost
gemini.google.com General conversation, quick tasks, everyday use Free (Advanced plan available)
Google AI Studio Prototyping, testing prompts, exploring models Free with generous limits
Gemini API Building applications, integration, production Free tier + pay-per-use
Google Workspace Docs, Sheets, Slides, Gmail integration Workspace subscription
Vertex AI Enterprise applications, fine-tuning, grounding Pay-per-use

Step 1: Your First Conversation

The easiest way to start is through the web interface:

  1. Visit gemini.google.com

    Go to gemini.google.com and sign in with your Google account.

  2. Start a conversation

    Type a message in the input box at the bottom. Try something like: "Explain how solar panels work in simple terms."

  3. Review the response

    Gemini will generate a response. You can ask follow-up questions in the same conversation to dive deeper into any topic.

  4. Try different features

    Notice the options to modify responses, share them, or start a new conversation. Explore the settings to switch between models if available.

Example Conversation 
You:
I'm planning a trip to Japan in April. What should
I know about cherry blossom season?

Gemini:
Cherry blossom season (sakura) in Japan typically
runs from late March to mid-April. Here's what you
should know:

Timing: Blossoms start in southern Kyushu and
move north. Tokyo usually peaks in late March to
early April. Kyoto follows about a week later.

Best Viewing: Popular spots include Ueno Park
(Tokyo), Philosopher's Path (Kyoto), and Osaka
Castle Park. Arrive early for the best spots.

Hanami: The tradition of flower viewing involves
picnics under the trees. Bring a tarp, snacks, and
drinks to join the celebrations...

Step 2: Using Images and Files

Gemini's multimodal capabilities let you work with more than just text:

Uploading Images

Click the image icon in the input area to upload a photo. Then ask Gemini about it:

Multimodal Prompt Examples 
# With a photo of a plant:
"What plant is this? Is it safe for cats?"

# With a screenshot of an error:
"What does this error mean and how do I fix it?"

# With a photo of a menu in another language:
"Translate this menu and recommend a dish."

# With an architecture diagram:
"Explain this system architecture and suggest improvements."

Working with Files

You can upload documents like PDFs, spreadsheets, and text files for Gemini to analyze:

  • PDFs: Upload reports, papers, or documents for summarization or Q&A
  • Code files: Share source code for review, debugging, or explanation
  • Spreadsheets: Upload data files for analysis and visualization suggestions
✅
Tip: When uploading images, provide specific questions rather than vague ones. "What breed is this dog and approximately how old does it look?" works better than just "Tell me about this image."

Step 3: Google AI Studio

For more control and experimentation, use Google AI Studio:

  1. Open AI Studio

    Visit aistudio.google.com and sign in with your Google account.

  2. Create a new prompt

    Choose between Freeform, Chat, or Structured prompts depending on your task.

  3. Configure settings

    Select a model (Pro, Flash, etc.), adjust temperature, set max output tokens, and add safety settings.

  4. Get your API key

    Click "Get API Key" to generate a key for programmatic access. This key works with the Gemini API.

💡
AI Studio is free: Google AI Studio provides generous free usage limits. It is the fastest way to prototype and test prompts before building them into applications via the API.

Step 4: Google Workspace Integration

If you have a Google Workspace plan with Gemini, you can use AI directly in your favorite tools:

📄

Google Docs

Use "Help me write" to draft, rewrite, or summarize content. Gemini can generate entire documents from a brief description.

📈

Google Sheets

Use "Help me organize" to generate formulas, create tables from descriptions, and analyze data patterns.

🎨

Google Slides

Use "Help me visualize" to generate slide content, create image placeholders, and design presentation outlines.

✉

Gmail

Use "Help me write" to draft emails, suggest replies, and summarize long email threads.

💡 Try It: First Multimodal Prompt

Go to gemini.google.com, upload a photo from your phone or desktop, and ask Gemini a specific question about it. Try asking for analysis, identification, or creative suggestions based on the image.

Compare Gemini's image understanding with a text-only prompt. Notice how the multimodal input changes the quality and specificity of the response.