Azure AI Foundry — no Marketplace label or sponsorship credit warning when deploying third-party models. Is this expected behavior?

Question

I'm a Microsoft for Startups participant and I want to understand whether the following behavior in Azure AI Foundry is expected or a UI gap.

When deploying AI models (specifically Anthropic Claude) through Azure AI Foundry, the deployment flow presents all models — both Microsoft-billed and third-party — in the same interface. There is:

• No "Marketplace" label visible during the deployment flow
• No warning that the model will be billed separately from Azure sponsorship credits
• No confirmation prompt indicating that credits will not apply
• No visual distinction between native Azure services and third-party Marketplace offers

As a Startups program participant with active sponsorship credits, the interface gave me the impression that credits would apply to all services deployed within the portal. I only learned that Marketplace charges were being billed separately after the fact, and no billing notification was sent.

I've seen this reported by other founders as well: https://learn.microsoft.com/en-us/answers/questions/5728641/azure-ai-foundry-presents-third-party-ai-models-as

My questions:

1. Is the absence of a Marketplace label or credit exclusion warning in the Azure AI Foundry deployment flow expected behavior, or is this a known UI gap?
2. Is there a way for Startups program participants to identify which models in AI Foundry are covered by sponsorship credits before deploying them?
3. Has Microsoft acknowledged this as a UX issue, and is a fix planned?

Any insight from Microsoft employees or community members who have encountered this would be helpful. Thank you.I'm a Microsoft for Startups participant and I want to understand whether the following behavior in Azure AI Foundry is expected or a UI gap.

When deploying AI models (specifically Anthropic Claude) through Azure AI Foundry, the deployment flow presents all models — both Microsoft-billed and third-party — in the same interface. There is:

• No "Marketplace" label visible during the deployment flow
• No warning that the model will be billed separately from Azure sponsorship credits
• No confirmation prompt indicating that credits will not apply
• No visual distinction between native Azure services and third-party Marketplace offers

As a Startups program participant with active sponsorship credits, the interface gave me the impression that credits would apply to all services deployed within the portal. I only learned that Marketplace charges were being billed separately after the fact, and no billing notification was sent.

I've seen this reported by other founders as well: https://learn.microsoft.com/en-us/answers/questions/5728641/azure-ai-foundry-presents-third-party-ai-models-as

My questions:

1. Is the absence of a Marketplace label or credit exclusion warning in the Azure AI Foundry deployment flow expected behavior, or is this a known UI gap?
2. Is there a way for Startups program participants to identify which models in AI Foundry are covered by sponsorship credits before deploying them?
3. Has Microsoft acknowledged this as a UX issue, and is a fix planned?

Any insight from Microsoft employees or community members who have encountered this would be helpful. Thank you.

Answer 1

Hey CZ – you’re correct that the current Foundry deployment UX doesn’t call out Marketplace-billed (third-party) models vs. Azure-native models, and that can lead to surprise billing. Here’s what you need to know:

1. Is this expected or a known UI gap? • Today Azure AI Foundry shows all models side-by-side in the same catalog and deployment flow, without a “Marketplace” badge or an explicit “your sponsorship credits won’t apply” warning. That is expected behavior in the current UI, although we’ve heard multiple pieces of feedback that it should be more clear.
2. How can Startups participants identify credit-eligible models up front? • Use the “Direct from Azure” collection in the Foundry catalog – models in that view are billed directly by Microsoft and will consume your sponsorship credits. • In the general catalog or via SDK, click into a model’s details and check the Publisher field: – Publisher = Microsoft → eligible for sponsorship credits – Publisher = Anthropic, OpenAI (non-Azure), or any third-party → billed separately via Marketplace • After deployment you can also validate actual billing sources in Azure Cost Management (Azure AI / Azure OpenAI usage vs. Marketplace usage).
3. Has Microsoft acknowledged this and is a fix planned? • Yes – the product team has this on their backlog as a UX improvement item. They’re evaluating ways to introduce badging or a confirmation prompt to call out non-Azure-native models. I don’t have an exact timeline, but this is active feedback.

In the meantime, I recommend always starting your model selection from the “Direct from Azure” filter or checking the Publisher field before you click Deploy. That will ensure your Microsoft for Startups sponsorship credits cover the usage.

Hope that helps!

Reference list

• Eligibility criteria for AI Foundry model sponsorship in Microsoft for Startups: https://learn.microsoft.com/microsoft-for-startups/build/foundry-model-sponsorship-coverage

Note: This content was drafted with the help of an AI system. Please verify the information before relying on it for decision-making.

Answer 2

1. The current behavior in Azure AI Foundry where third-party models (for example, Anthropic Claude via Marketplace) appear in the same deployment experience as Microsoft-billed models is consistent with how Foundry exposes the broader Azure AI model catalog. Documentation emphasizes that the catalog includes both:
   • Models sold directly by Azure (first-party)
   • Models from partners and the community (third-party), often billed via Azure Marketplace or separate agreements
   Azure credits and Azure Sponsorship explicitly do not cover third-party/Marketplace offerings, even when they are surfaced in the same portal experience. This is documented as a billing and sponsorship rule rather than as a UI behavior guarantee. The provided context does not describe a specific, documented requirement that the Foundry UI must show a “Marketplace” label or sponsorship warning during deployment, nor does it describe this as a known UI bug or gap.
2. For Microsoft for Startups participants, the supported way to determine whether a model is covered by sponsorship credits is to check the publisher and billing source, not just the portal surface:
   • Sponsorship credits apply only to Microsoft-billed AI models in Azure AI Foundry.
   • Models billed by third-party providers or via Azure Marketplace are not eligible.
   The documented steps to verify eligibility are:
   1. In Azure AI Foundry, open the model and check the Publisher field:
      • Publisher = Microsoft → eligible for Microsoft for Startups sponsorship credits
      • Publisher = non-Microsoft → not eligible
   2. Optionally confirm in Azure Cost Management:
      • Charges under Azure AI / Azure OpenAI usage → eligible
      • Charges under Marketplace or third-party usage → not eligible
   Additionally, the “Direct from Azure” collection in Azure AI Foundry is called out as a discovery aid:
   • Models in this collection are “commonly eligible” for Microsoft for Startups sponsorship credits because they are typically hosted and billed directly by Microsoft.
   • Eligibility should still be confirmed using the publisher/billing checks above.
   More generally, Azure Sponsorship and Azure credits have the following constraints that apply to startups:
   • Azure credits can be used for first-party Azure products, including Azure OpenAI and other first-party Azure AI models, when billed as normal Azure consumption on the sponsored subscription.
   • Azure credits and Azure Sponsorship do not cover:
     • Third-party (3PP) products or services, including partner-provided offerings and non–first-party Azure Marketplace products
     • Models from partners and the community in Foundry that are billed via Marketplace or separate third-party agreements
3. The documentation in scope acknowledges the issue at the policy and billing level, not as a formally documented UX defect with a published remediation plan:
   • Microsoft for Startups guidance explicitly distinguishes:
     • First-party Azure AI models deployed via Azure AI Foundry on a sponsored subscription → automatically covered by Azure credits
     • Third-party AI models → not automatically covered; eligibility varies and must be confirmed against sponsorship rules
   • Azure Sponsorship documentation explicitly clarifies that:
     • Models sold directly by Azure (1PP) may be covered by sponsorship credits
     • Models from partners and the community (3PP) are not covered
   The available context does not state that Microsoft has officially documented the lack of Marketplace/credit warnings in the Foundry UI as a UX issue, nor does it describe any committed fix or roadmap item. The official guidance instead focuses on:
   • Using the publisher field and billing source to determine eligibility
   • Using collections like “Direct from Azure” as a shortcut for commonly eligible models
   • Relying on Azure Cost Management to see how usage is billed (Azure-native vs Marketplace/third-party)

In practice, for Microsoft for Startups participants, the safe workflow before deploying a model from Azure AI Foundry is:

1. Open the model in Azure AI Foundry and confirm Publisher = Microsoft.
2. Prefer models in the “Direct from Azure” collection when possible.
3. For any non-Microsoft publisher or Marketplace-based model, assume sponsorship credits will not apply and that charges will be billed separately.
4. After initial usage, validate in Azure Cost Management that charges appear under Azure AI / Azure OpenAI (eligible) and not under Marketplace/third-party (ineligible).

---

References:

• Eligibility criteria for AI Foundry model sponsorship in Microsoft for Startups
• How to use Azure credits for GitHub, AKS, and AI Models
• Track Microsoft Customer Agreement Azure credit balance
• Azure experience for founders
• Deploy Microsoft Foundry Models to managed compute with pay-as-you-go billing (classic)
• Foundry Models frequently asked questions