Azure AI Foundry & Vertex AI for AI Agents

Azure AI Foundry & Google Vertex AIATX Agent Hub is multi-model by design. Each AI agent workflow chooses its model from Azure AI Foundry or Google Vertex AI Model Garden, and a comparison step can pick the best answer across providers — so a new model is a setting, not a rebuild.

What you can do

Azure AI Foundry: OpenAI GPT, Claude, Llama, Mistral, DeepSeek, Cohere and more
Google Vertex AI Model Garden: Gemini, Imagen, Veo, and partner and open models
Model Comparison: run the same step on several models and select the best answer
Embedding models for retrieval-augmented generation over your own knowledge bases
Token and cost tracking per project, with budget caps and alerts

How to connect Azure AI Foundry & Google Vertex AI

An administrator configures Azure AI Foundry and Google Vertex AI access once, with credentials in the vault.
On the workflow canvas, add an AI Model node and choose the model for that step.
Add Model Comparison to try several models on the same step, and watch token cost per project.

Security and governance

API keys and service credentials live in the encrypted credential vault — AES-256 at rest, scoped per project, never shown in logs
AI Guardian, Fact Checking, Intent Check and Domain Constraint nodes check model output before it is used

Related guides

More integrations

Glossary terms on this page

ATX Institute on LinkedIn

This site needs JavaScript for the full experience. Every page above is reachable without it.