azure-ai-projects

What it is and what it does

The azure-ai-projects package is the Python client for Microsoft Foundry, a managed AI platform on Azure. It provides programmatic access to Foundry projects, allowing you to create and run agents (including containerized Hosted Agents), attach specialized tools and integrations, manage toolboxes, datasets, search indexes, and AI model deployments. The package wraps the Foundry v1 REST API and includes both stable and preview features; preview capabilities require setting allow_preview=True or accessing them through .beta sub-clients.

Typically used by developers building AI applications on Azure who need to orchestrate agents, manage resources, and integrate with Azure services like Storage, Search, Functions, and Fabric. The package depends on azure-core for HTTP handling, azure-identity for Entra ID authentication, openai for LLM operations, and azure-storage-blob for document management. It supports synchronous and asynchronous workflows and is actively maintained with support for Python 3.9 through 3.14.

Use it for:

• Build and deploy multi-agent systems that leverage Azure's managed hosting and scaling infrastructure.
• Create agents with specialized tools (search, code execution, web browsing, image generation) without managing infrastructure.
• Fine-tune AI models on your own data and manage model deployments within a Foundry project.
• Manage datasets, search indexes, and document uploads for retrieval-augmented generation (RAG) workflows.
• Integrate agents with external systems via OpenAPI, Azure Functions, and Microsoft Fabric for enterprise automation.
• Evaluate and monitor generative AI application performance using built-in evaluation rules and red-team scanning.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Provides a Python client for Microsoft Foundry projects, enabling creation and management of AI agents, toolboxes, datasets, and integration with Azure AI services and OpenAI models.

Yes. The package is actively maintained, has low install friction, carries a permissive MIT license, and provides a comprehensive, well-documented interface to Microsoft Foundry. It is suitable for production use of stable features. If you are building AI agents on Azure or need to manage Foundry resources programmatically, this is the standard choice. Be aware that preview features are subject to change and should not be used in production without careful testing.

Install

azure-ai-projects on PyPI

pip install azure-ai-projects

uv add azure-ai-projects

poetry add azure-ai-projects

Installing azure-ai-projects

Before you install

Low installation friction with a pure-Python wheel. Actively maintained with a release 18 days ago. Depends on established Azure SDK packages (azure-core, azure-identity, azure-storage-blob) and openai, all standard in the ecosystem.

License in practice

MIT license permits unrestricted use, modification, and distribution with minimal restrictions—suitable for both commercial and open-source projects.

Quickstart

pip install azure-ai-projects

import os
from azure.ai.projects import AIProjectClient
from azure.identity import DefaultAzureCredential

with DefaultAzureCredential() as credential:
    client = AIProjectClient(
        endpoint=os.environ["FOUNDRY_PROJECT_ENDPOINT"],
        credential=credential
    )

Requires Python 3.9 or later, an Azure subscription, a Microsoft Foundry project, and Entra ID authentication via DefaultAzureCredential (typically requires Azure CLI and az login).

Verify before relying

• Whether async support requires aiohttp as an optional dependency or if it is bundled.
• Performance characteristics when managing large numbers of agents or datasets.
• Stability guarantees for preview features marked as such in the API.

Package facts

Evidence: azure_ai_projects-2.4.0-py3-none-any.whl

Keywords: azure, azure sdk

Tags

More Artificial Intelligence packages

Further reading