Skip to main content
Microsoft Foundry−26 Negative

GILD Review

GILD's call and the buying facts behind it.

From GILD research

Should we adopt Microsoft Foundry?

What does a team get from building on Microsoft Foundry compared with buying each model from its own vendor?

  1. 30-Second TakePick Microsoft Foundry if you want OpenAI, Anthropic and open models under Entra ID and one Azure agreement.
  2. Best fitTeams on Azure buying OpenAI and Anthropic models under one agreement and one identity system.GILD Review
  3. Consider another workflowIf you need Gemini, or Claude features Anthropic ships first, buy from Google Cloud or Anthropic.GILD Review

Drawn from 59 sources and 4 operator insights

GILD analysis

Microsoft Foundry Review

Foundry puts OpenAI's models, Claude and open models inside an Azure subscription under Entra ID, with Global, Data Zone and regional deployments and Foundry Agent Service for agents. The cost is uneven terms: quota tiers follow usage and agreement status, and Claude needs a paid pay-as-you-go subscription. Requesty's gateway measured 78% request success on Azure in April 2026.

  • AI Model APIs
  • AI Costs

The decision

Best fit
Teams on Azure buying OpenAI and Anthropic models under one agreement and one identity system.
Consider another workflow
If you need Gemini, or Claude features Anthropic ships first, buy from Google Cloud or Anthropic.
Adoption condition
Confirm the quota tier and deployment type for each model, and set a fallback route, before production traffic.

Basis: Microsoft Learn checked from Oct 28 to Nov 1, 2026, Azure pricing and the Azure retail prices API checked Oct 28, 2026, Microsoft's Sweden Central post-incident review, and public practitioner accounts. Requesty's gateway data supplies the success rate. No GILD performance benchmark.

Buying & deployment factsWhat a team building on it should verify before production traffic grows.
How you buy it

US$2 in, US$10 out

GPT-6.1 Sol per million tokens on Global Standard in East US 2, from Oct 1, 2026.

Source 1, prices.azure.com, Retail prices, GPT-6 meters: Retail prices, GPT-6 metersSource 2, learn.microsoft.com, Deployment types: Deployment types

Details

Standard deployments bill per million tokens: GPT-6.1 Sol on Global Standard in East US 2 lists US$2 input and US$10 output per million tokens, effective Oct 1, 2026. Provisioned deployment types bill for reserved throughput units (PTU) and carry no per-token charge.

Cost drivers

Batch at 50% off

Global Standard is the lowest price; provisioned types reserve throughput for steady volume.

Source 2, learn.microsoft.com, Deployment types: Deployment typesSource 5, Microsoft, Microsoft Learn: Model router for Microsoft Foundry concepts: Microsoft Learn: Model router for Microsoft Foundry concepts

Details

Deployment type sets much of the bill. Global Standard has the lowest price, and Global Batch and Data Zone Batch take 50% off for large asynchronous jobs, Global Batch within a 24-hour window. Provisioned types reserve throughput for predictable high volume. Model router routes each prompt in real time to the most suitable model, on Global Standard and Data Zone Standard deployments.

Where it runs

Global, zone or region

Global, Data Zone (US, EU or Asia Pacific), or regional processing, set per deployment.

Source 2, learn.microsoft.com, Deployment types: Deployment typesSource 6, learn.microsoft.com, Microsoft Learn: Use Microsoft Purview to manage data security & compliance for Microsoft Foundry: Microsoft Learn: Use Microsoft Purview to manage data security & compliance for Microsoft FoundrySource 7, learn.microsoft.com, Microsoft Learn: ISO/IEC 42001:2023 compliance offering: Microsoft Learn: ISO/IEC 42001:2023 compliance offeringSource 4, Microsoft, Data, privacy, and security: Data, privacy, and security

Details

Global deployments may process in any Azure region, Data Zone deployments only inside the US, EU or Asia Pacific zone, and Standard or regional provisioned deployments inside the customer-specified Azure geography. Data stored at rest stays in the designated Azure geography. With Microsoft Purview Data Security enabled, Foundry prompts and responses are captured in the unified audit log. Purview retention policies can retain or delete those prompts and responses. Microsoft lists Microsoft Foundry in its ISO/IEC 42001 certification scope.

What to verify

No SSO or SCIM documented

0 matches on 1,350 Microsoft Learn pages (Oct 29, 2026).

Source 3, learn.microsoft.com, Quotas and limits: Quotas and limitsSource 4, Microsoft, Data, privacy, and security: Data, privacy, and security

Details

Quota comes in seven tiers, set per subscription. Microsoft documents no SSO, SAML or SCIM for Foundry: GILD found 0 matches across the 1,350 Foundry pages on Microsoft Learn, the Product Terms and the Trust Center, checked Oct 29, 2026.

Pages GILD searched, by set, Oct 29, 2026
WherePages readMatches
Microsoft Learn: Foundry, Foundry (classic) and Foundry Local docs, including Azure OpenAI1,3500
Microsoft Learn: security baselines, authentication and the Foundry reference architecture70
Microsoft Product Terms, Microsoft Azure10
Microsoft Trust Center and Azure compliance pages40
Azure product pages for Foundry, Azure OpenAI and Foundry Agent Service30
Microsoft Foundry blog: all 111 posts, read in 2 feed pages20
GILD search record, Oct 29, 2026

Sourced from Microsoft documentation