Microsoft Foundry (New) vs Foundry (Classic) — What Actually Changed?

I'm a software consultant who loves to build projects and share my learnings on this blog.
Search for a command to run...

I'm a software consultant who loves to build projects and share my learnings on this blog.
No comments yet. Be the first to comment.
Azure Functions is a powerful serverless compute platform, and Azure App Service handles a lot of the underlying configuration for you out of the box. However, there are scenarios where you may want t

In today's digital age, speed is paramount for web and mobile applications. Users expect instantaneous responses, and any delay can lead to a poor user experience and loss of engagement. Therefore, as developers, it's crucial to build web APIs that a...

Loops are fundamental constructs in any programming language. But have you ever wondered if there are any performance differences between them? For small collections, the difference might be negligible, but what about handling a million or 10 million...

When using an Azure Function app, it is necessary to connect it to a storage account, often referred to as Host storage. This storage is utilized by the Functions runtime, as well as by various triggers and bindings to coordinate between multiple run...

If you’ve been exploring Azure AI learning paths or watching older demos, you’ve probably run into something confusing:
The documentation and training links often reference Foundry (Classic)
But the portal you’re using today might show Foundry (New)
And the way you create projects and resources looks slightly different
This blog is a practical guide on what changed, how “Classic” and “New” relate, and what you’ll actually see in your Azure subscription.
At a high level, Microsoft Foundry is Azure’s unified platform for building, deploying and governing modern AI applications—especially those built around agents, AI workflows, evaluations, tools, and observability.
Instead of stitching together isolated services for model hosting, agent orchestration, monitoring, and governance, Foundry brings all of these capabilities into a single experience.
Think of it as Azure’s “all-in-one” platform for AI development.
Before comparing Classic and New Foundry portals, it’s important to get one thing right:
Foundry has two project types.
Hub-based Projects (older model)
Foundry Projects (modern unified model)
Everything else—Classic vs New portals—is essentially different UX layers sitting on top of these project types.
This is the pre-Foundry architecture, built around:
Azure AI Hub
Azure AI Project
Supporting resources (storage accounts, managed identities, network components, etc.)
This was the default model before Foundry unified the platform.
As a result, Foundry (Classic) continued to support and create these hub-based setups.
When you create a hub-based project, Azure provisions:
Azure AI Hub resource
Azure AI Project resource
You’ll still see references such as:
Azure AI Hub
AI Project
Project under a hub
Most older tutorials, Prompt Flow documentation, and Learn modules are based on this model. Also, Prompt Flow is currently supported only in hub-based projects.
Foundry Classic = supporting the old Hub-based architecture.

This is the recommended, simplified, and future-facing project type.
A Foundry project includes:
Foundry resource
Foundry Project resource
There is no hub in this model.
This architecture simplifies:
Access control
Monitoring & observability
Resource management
Agent workflows
Tooling integrations
Lifecycle operations
When you create a Foundry project, Azure provisions:
Foundry resource
Foundry Project resource
Foundry New = hub-less, modern, and the future of the platform.

There are two Foundry portal experiences, but both operate on the same platform.
This is the older UI, and most existing training content still references it.
Lets you create both Hub-based Projects and Foundry Projects
Still widely used because of older documentation
Provides the UI that hub-based resources depend on

This is the modern experience with a cleaner UI and a project model aligned to the future of Azure AI development.
Only allows creation of Foundry Projects
Streamlined navigation
Better integrations for agents, tools, workflows and evaluations
A unified structure without hubs

In the Foundry portal, you can switch between Classic and New using the toggle at the top:

The official docs also include a Version toggle for Classic vs New—very useful when your UI doesn’t match the tutorial:

Now that we’ve seen both portals, here’s a simple comparison to keep in mind:
| Area | Foundry (Classic) | Foundry (New) |
| UI model | Older portal experience | Modern, simplified UI |
| Supports Hub-based Projects | ✅ Yes | ❌ No |
| Supports Foundry Projects | ✅ Yes | ✅ Yes |
| Recommended for new apps | ❌ No | ✅ Yes |
| Prompt Flow availability | Only in Hub-based Projects | Not yet supported |
| Architecture focus | Hub + AI Project | Foundry + Foundry Project |
When in doubt:
If you’re starting something new, use Foundry (New) with a Foundry Project.
If you’re following an older tutorial or managing an existing hub-based environment, you may still need Foundry (Classic).
If your team is still running on hub-based projects, Microsoft supports migrating to the modern Foundry Project model. The migration involves moving:
Model deployments
Data connections
Tools & workflows
Identity & access policies
Depending on how the old project was structured, some steps may be manual.
If a tutorial shows Classic, just switch the docs to the matching version.
If you’re creating something new, always choose a Foundry Project in the new portal.
Prompt flow is currently supported only in Hub based projects.
Expect naming confusion initially—AI Hub, Foundry, AI Project, Foundry Project—because the platform has been evolving.
The difference between Foundry (New) and Foundry (Classic) is not about two different products.
It’s about one platform moving from an older Hub-based architecture to a cleaner, unified project model.
To keep it simple in your mind:
Foundry (New) + Foundry Project
→ Modern, streamlined, and the recommended default for all new scenarios
Foundry (Classic) + Hub-based Projects
→ Older architecture, still supported, primarily used when following legacy tutorials or working with older deployments
As the Azure AI ecosystem continues to evolve, the Foundry Project model is where Microsoft is investing—so that’s where you’ll want to be.