Migrate Azure AI Content Understanding from GPT-4.1 to GPT-5.2

Content Understanding supports GPT-5.2 as a chat completion model, but Studio currently highlights GPT-4.1 and GPT-4.1-mini in the model picker. GPT-5.2 is recommended and can be used by configuring model deployments and defaults at the resource/analyzer level rather than relying solely on the visible Studio dropdown.

To migrate from GPT-4.1 to GPT-5.2 in Azure AI Content Understanding:

1. Deploy GPT-5.2 in Foundry
   • In the Foundry resource, create a model deployment for gpt-5.2 (chat completion). Content Understanding connects to Foundry models for generative capabilities and lists gpt-5.2 as a supported chat completion model.
2. Set GPT-5.2 as the default completion model
   • Use the model deployment options described for Content Understanding:
     • Either set default model deployments at the resource level (via PATCH /contentunderstanding/defaults in the REST API), or
     • Pass the gpt-5.2 deployment in each modelDeployments block of analyze requests.
   • Once defaults are set, analyze requests can omit modelDeployments and will use the configured defaults.
3. Bind analyzers to the GPT-5.2 alias
   • When defining or updating a custom analyzer, set its models.completion to the alias that maps to the gpt-5.2 deployment (for example, the same alias previously mapped to gpt-4.1).
   • GPT-5.2 is explicitly recommended as the primary model for Foundry and Studio, so using this alias ensures analyzers run on GPT-5.2 instead of GPT-4.1.
4. Use Studio with resource defaults
   • Studio can configure defaults for supported models such as gpt-5.2, gpt-4.1, gpt-4.1-mini, and text-embedding-3-large when a Foundry resource is connected and auto-deployment is enabled.
   • If Studio’s dropdown still shows only GPT-4.1 family, rely on the resource-level defaults and analyzer model aliases to route traffic to the gpt-5.2 deployment. The retirement notice for GPT-4.1 family (retiring October 2026) is already reflected in the service guidance, which recommends migrating to gpt-5.2.

If Studio UI does not yet expose GPT-5.2 in the picker for a specific region or tenant, the supported and documented path is to configure GPT-5.2 via Foundry deployments and Content Understanding defaults/aliases, then continue using Studio and APIs as usual.

References:

• Azure Content Understanding in Foundry Tools service quotas and limits
• Model deployment options for Content Understanding analyzers
• Migrate from Azure Content Understanding Preview to GA
• What's new in Azure Content Understanding in Foundry Tools?
• Frequently asked questions