Generative AI platforms — Azure OpenAI · AWS Bedrock · GCP Gemini
Generative AI platforms provide foundational LLMs as a service. Each cloud has its own offering with varying models and capabilities.
import vertexai
from vertexai.generative_models import GenerativeModel, Part
vertexai.init(project="my-project", location="us-central1")
model = GenerativeModel("gemini-2.5-pro")
# Text generation
response = model.generate_content(
"Explain how VPC Service Controls prevent data exfiltration."
)
print(response.text)
# Multimodal (image + text)
image = Part.from_uri("gs://my-bucket/diagram.png", mime_type="image/png")
response = model.generate_content(
[image, "Redraw this architecture as Terraform HCL."]
)Google AI Studio (aistudio.google.com) is a free web-based playground for prototyping Gemini prompts. Features: system instructions, temperature tuning, safety settings, token counting, and one-click "Get code" for Python/Node.js/curl.
Models: GPT-4o, GPT-4, o1. Content filtering, private networking, RBAC.
CLI: az cognitiveservices account create --kind OpenAI --name my-openai -g prod-rg --sku S0
Models: Claude 3, Llama 3, Titan, Mistral, Cohere. Serverless API. Model customization.
CLI: aws bedrock list-foundation-models
Models: Gemini 2.5 Pro (2M token context), Gemini Flash, Imagen. Integrated with Vertex AI for MLOps and governance.