Foundry DPA

Question

For FLUX.2-pro accessed via the BFL service provider API endpoint (/providers/blackforestlabs/v1/flux-2-pro), does inference run entirely on Microsoft-managed Azure infrastructure, or are requests routed to Black Forest Labs' servers?

And does the Microsoft Products and Services Data Protection Addendum apply fully to this model?

Answer 1

Hello Anthony,

Thank you for your question.

For FLUX.2-pro accessed via the provider endpoint:

/providers/blackforestlabs/v1/flux-2-pro

please find the clarifications below.

Inference location

Inference for this model is executed on Microsoft-managed Azure infrastructure.

• Azure AI Foundry serverless APIs run workloads within the selected Azure region
• Partner models, including FLUX.2-pro, are deployed and executed in containers hosted on Azure compute
• Requests are not routed to external Black Forest Labs infrastructure for inference

Data Protection Addendum (DPA)

Data processed through this endpoint is handled under the Microsoft Products and Services Data Protection Addendum (DPA).

• Input and output data are processed within Azure
• Microsoft’s standard data protection, security, and compliance commitments apply to the service operation

Partner model consideration

FLUX.2-pro is categorized as a non-Microsoft (partner) model.

• While execution and data processing occur within Azure, the model itself is provided by Black Forest Labs
• As such, provider-specific terms may apply for aspects such as model behavior, logging, or usage policies

Inference runs on Microsoft-managed Azure infrastructure

Requests are not forwarded to Black Forest Labs servers

The Microsoft DPA applies to data processed by the service

As a partner model, additional provider terms should be reviewed.

Please refer this

Deploy via Serverless APIs (Models-as-a-Service) in Azure AI Foundry – details on Azure-managed serverless compute https://learn.microsoft.com/azure/ai-foundry/how-to/deploy-models-serverless

Foundry models from partners and community – notes on non-Microsoft products & terms https://learn.microsoft.com/azure/foundry/foundry-models/concepts/models-from-partners?wt.mc_id=knowledgesearch_inproduct_azure-cxp-community-insider#note

Microsoft Products and Services Data Protection Addendum overview https://www.microsoft.com/licensing/docs/view/Microsoft-Products-and-Services-Data-Protection-Addendum

I Hope this helps. Do let me know if you have any further queries.

---

If this answers your query, please do click Accept Answer and Yes for was this answer helpful.

Thank you!

Answer 2

Hello Anthony,

Greetings!

Thanks for raising this question in Q&A forum.

These are two really important questions around infrastructure and data protection, and I'm glad you asked. Let me address both clearly.

Q1: Where does inference actually run for FLUX.2-pro via the BFL provider API endpoint?

The short answer is: inference runs on Microsoft-managed Azure infrastructure, not on Black Forest Labs' own servers.

Foundry Models are fully hosted and managed on Azure. Even though the endpoint path includes /providers/blackforestlabs/, this is simply how the BFL-specific API is routed within Azure — the actual compute stays within Azure's infrastructure.

BFL's collaboration with Microsoft started from their earliest days, when they partnered with Azure to build their training and inference clusters. The models are served from Azure — your requests do not leave Microsoft's environment and go to Black Forest Labs' own servers.

Q2: Does the Microsoft Products and Services Data Protection Addendum (DPA) apply to FLUX.2-pro?

Yes, it does. Here's the key point:

Black Forest Labs models in Foundry Models are delivered under the Microsoft Product Terms, giving you enterprise-grade security and compliance out of the box.

Additionally, from Microsoft's official data privacy documentation: Microsoft acts as the data processor for prompts and outputs sent to, and generated by, a model deployed for serverless API deployment. Microsoft doesn't share these prompts and outputs with the model provider. Also, Microsoft doesn't use these prompts and outputs to train or improve Microsoft models, the model provider's models, or any third party's models.

So to summarize both points clearly:

1. Your inference requests stay on Azure infrastructure they are not routed to Black Forest Labs' own servers.
2. The Microsoft DPA applies fully. Your prompts and outputs are not shared with BFL or any third party.

One thing to note: Foundry doesn't provide built-in content filtering for FLUX models at deployment time, so you'll want to configure Azure AI Content Safety separately if that's a concern for your use case.

If this answer helps you kindly accept the answer which will help others who have similar questions.

Best Regards,

Jerald Felix.