Gemini API & Google AI Studio

📖 3 min read deepmindgooglegeminiapisdkdevelopmentprovider-api

Complete guide to the Gemini API — AI Studio (free developer IDE), Vertex AI (enterprise), Gemini Live API (real-time voice), function calling, streaming, embeddings, and pricing.

Key Takeaways

Google AI Studio is the free developer IDE — no credit card required, 1M token context, access to the full Gemini 3 lineup (3.1 Pro preview, 3.5 Flash, 3.1 Flash-Lite)
Gemini API: text, image, audio, video input. Function calling, streaming, structured output
Vertex AI: enterprise deployment with Google Cloud integration, VPC, IAM, compliance
Gemini Live API: real-time bidirectional voice and audio streaming

Getting Started — Google AI Studio (Free)

Google AI Studio is the fastest way to start building with Gemini — no credit card required.

Go to aistudio.google.com
Sign in with your Google account
Start prompting — free tier includes Gemini 3.5 Pro access

# Python SDK
pip install google-generativeai

import google.generativeai as genai
genai.configure(api_key="YOUR_API_KEY")

model = genai.GenerativeModel("gemini-3.5-pro")
response = model.generate_content("Explain quantum computing in one paragraph")
print(response.text)

Gemini API

Text Generation

model = genai.GenerativeModel("gemini-3.5-flash")
response = model.generate_content("Write a Python function to check if a number is prime")
print(response.text)

Multimodal Input

import PIL.Image

# Text + image
image = PIL.Image.open("photo.jpg")
model = genai.GenerativeModel("gemini-3.5-pro")
response = model.generate_content(["Describe this image in detail:", image])
print(response.text)

Audio Input

# Audio file analysis
audio_file = genai.upload_file("meeting.mp3")
model = genai.GenerativeModel("gemini-3.5-pro")
response = model.generate_content(["Summarize this meeting:", audio_file])
print(response.text)

Video Input

# Video analysis
video_file = genai.upload_file("presentation.mp4")
model = genai.GenerativeModel("gemini-3.5-pro")
response = model.generate_content([
    "Analyze this presentation and create a summary:", video_file
])
print(response.text)

Streaming

model = genai.GenerativeModel("gemini-3.5-flash")
response = model.generate_content("Write a haiku about programming", stream=True)
for chunk in response:
    print(chunk.text, end="", flush=True)

Function Calling

model = genai.GenerativeModel(
    "gemini-3.5-pro",
    tools=[{
        "function_declarations": [{
            "name": "get_weather",
            "description": "Get current weather for a location",
            "parameters": {
                "type": "object",
                "properties": {
                    "location": {"type": "string", "description": "City and state"}
                },
                "required": ["location"]
            }
        }]
    }]
)

Structured Output

model = genai.GenerativeModel("gemini-3.5-pro")
response = model.generate_content(
    "Extract people from: Alice met Bob at 3 PM",
    generation_config={"response_mime_type": "application/json"}
)

Gemini Live API — Real-Time Voice

Bidirectional streaming for voice agents and real-time audio:

# Client-side WebSocket connection for live voice
# Available via AI Studio or Vertex AI
# Supports: voice input → model processing → voice output
# Use cases: voice agents, live translation, audio analysis

Vertex AI — Enterprise

from google.cloud import aiplatform
from vertexai.generative_models import GenerativeModel

# Deploy within your GCP project
model = GenerativeModel("gemini-3.5-pro")
# Full GCP integration: IAM, VPC, Cloud Logging, audit trails

Feature	AI Studio	Vertex AI
Cost	Free tier, then pay-per-use	Pay-per-use, committed use discounts
Setup	No setup required	GCP project + APIs
Security	Google account	IAM, VPC, Private Link, Cloud KMS
Compliance	Standard	SOC 2, HIPAA, FedRAMP
Context	1M tokens	1M tokens
SLAs	None	Enterprise SLAs

Embeddings

model = "models/text-embedding-004"
result = genai.embed_content(
    model=model,
    content="The quick brown fox jumps over the lazy dog"
)
vector = result['embedding']

# Dimensions: 768 (text-embedding-004)
# Use for: semantic search, clustering, recommendations

Pricing

Gemini API pricing is tiered. Check ai.google.dev/pricing for latest:

Model Tier	Typical Range	Notes
Flash tier	$0.075-$ 0.15 / 1M input	Fastest, most affordable
Pro tier	$1.25-$ 2.50 / 1M input	Balanced, production workloads
Ultra tier	Premium pricing	Maximum quality
Audio/Live	Per-minute + per-token	Real-time streaming

AI Studio free tier: Generous free quota for experimentation. No credit card needed.

Where Next

Antigravity & Flow — agentic development platform
Media & Creative — video, image, music generation
Enterprise & Deployment — Vertex AI, compliance