An Azure service that provides access to OpenAI’s GPT-3 models with enterprise capabilities.
gpt-5.4-mini is currently available as a Global Standard model in specific regions (East US2, Sweden Central, South Central US, Poland Central) for Azure OpenAI, but there is no documentation indicating support for gpt-5.4-mini in Foundry Agent Service (agents) yet.
The Agents (classic) model matrix lists up to gpt-5 / gpt-5-mini and earlier GPT‑4.x/4o variants, but does not include gpt-5.4-mini as an agent-supported model. The reasoning models article confirms gpt-5.4-mini availability for Global Standard usage, but does not extend that to agents or provide any timeline for when it will be usable in agents.
Model and feature availability for agents is centrally managed by Microsoft and exposed via the documented region/model tables; there is no supported way to request enablement of a specific new model (like gpt-5.4-mini) for agents ahead of the published roadmap, and no ETA is published in the provided documentation.
At this time, only the models explicitly listed as supported for Foundry Agent Service can be used to create agents. For newer models such as gpt-5.4-mini, usage is limited to the supported deployment types (for example, Global Standard or provisioned throughput) shown in the model availability tables, not agents.
References: