What it's used for

Google Gemini is Google's frontier model family supporting multimodal inputs spanning text, images, audio, video, and code with extremely long context windows up to 1 million tokens. It is the most versatile multimodal model available, able to process entire video files, codebases, and large document collections in a single prompt.

Key use cases include:

Multimodal understanding — analyzing images, PDFs, audio, and video natively
Large-scale document processing with the 1M-token context window
Code generation and debugging across multiple programming languages
Grounded responses using Google Search integration
Building apps integrated with Google Cloud services
Rapid prototyping in Google AI Studio

Gemini is used by developers building applications that need deep integration with the Google ecosystem, multimodal capabilities, or ultra-long context. The model family includes Gemini 2.5 Pro (frontier), Gemini 2.5 Flash (balanced), and Gemini 2.0 Flash Lite (fast/cheap).

Google offers both a free-tier through AI Studio for prototyping and production-grade access through Vertex AI on Google Cloud.

Getting started

Get a free API key from Google AI Studio for prototyping, or set up Vertex AI in Google Cloud for production.

Install the SDK:

# Python
pip install google-genai

# Node.js
npm install @google/genai

Set your API key:
```
export GOOGLE_API_KEY='AIza...'
```

Make your first call:

from google import genai
client = genai.Client()
response = client.models.generate_content(
    model='gemini-2.5-flash',
    contents='Hello!'
)

Pricing: Free tier in AI Studio with generous rate limits. Gemini 2.5 Flash is ~$0.15/1M input tokens on the paid tier. Gemini 2.5 Pro is ~$1.25/1M input tokens. See ai.google.dev/pricing for current rates.