{"title":"AI models for agents (preview)","url":"https://jonas-paprotka.de/docs/dev-itpro/ai/ai-agent-models/","by":"JOP-AL","license":"Copyright (c) Microsoft Corporation. CC BY 4.0, https://creativecommons.org/licenses/by/4.0/; code samples MIT. From https://learn.microsoft.com/en-us/dynamics365/business-central/dev-itpro/ai/ai-agent-models, MicrosoftDocs/dynamics365smb-devitpro-pb at 828b6ee (2026-11-19). Changed: includes filled in, links to this site, and for agents shortened without dropping a fact.","credit":"Microsoft Corporation, CC BY 4.0","origin":"https://learn.microsoft.com/en-us/dynamics365/business-central/dev-itpro/ai/ai-agent-models","source":{"name":"Business Central developer and IT pro docs","repo":"https://github.com/MicrosoftDocs/dynamics365smb-devitpro-pb","commit":"828b6ee72fcd75755627cd6011f3d20899217ab2","date":"2026-11-19"},"checked":"2026-11-19","description":"Learn about AI model availability, lifecycle, and how to select models for agents in Dynamics 365 Business Central.","date":"2026-06-02","meta":{"topic":"conceptual"},"outline":[{"level":1,"title":"AI models for agents (preview)","slug":"ai-models-for-agents-preview","at":0,"chars":4916},{"level":2,"title":"Where to see and control models","slug":"where-to-see-and-control-models","at":1228,"chars":493},{"level":2,"title":"Select a model for a designed agent","slug":"select-a-model-for-a-designed-agent","at":1721,"chars":1053},{"level":2,"title":"Model resolution","slug":"model-resolution","at":2774,"chars":519},{"level":2,"title":"Model lifecycle","slug":"model-lifecycle","at":3293,"chars":874},{"level":2,"title":"Retired and obsolete models","slug":"retired-and-obsolete-models","at":4167,"chars":390},{"level":2,"title":"Related information","slug":"related-information","at":4557,"chars":359}],"text":"# AI models for agents (preview)\n[This article is prerelease documentation and is subject to change.]\n> IMPORTANT\n > This is a preview feature. Preview features might have restricted functionality and are subject to supplemental terms of use<https://www.microsoft.com/en-us/business-applications/legal/supp-powerplatform-preview/>.\n > The AI development toolkit is available in sandbox environments only. The Agent SDK for AL is available in sandbox environments and, starting from version 28.1, also in production environments.\n > This article is primarily intended for product owners, domain experts, consultants, and developers who want to experiment with AI capabilities in Business Central.\nAgents in Business Central can run on different AI models. The model an agent uses can affect its behavior, performance, and results. Multiple models can be available at the same time, and each agent can either use a specific model or rely on the automatic (default) model selection.\n> NOTE\n> The AI models available for agents in Business Central are determined by Microsoft and may change over time. Model availability can depend on factors such as the agent's capabilities, the environment, and the stage of the model lifecycle.\n## Where to see and control models\nYou can view and manage AI models in two ways:\n Agent configuration in the UI\n  For agents created from the agent design experience, the agent card and setup pages show the current model and allow you to change it. A dedicated list page shows all available models. For coded agents, the model in use isn't visible in the UI.\n Agents coded in AL\n  APIs allow you to list available models and set the model on an agent or on individual tasks programmatically.\n## Select a model for a designed agent\nTo change the AI model for an agent created in the design experience:\n1. In Business Central, go to the Agents page.\n2. Select the agent you want to configure.\n3. On the agent card, find the Agent model section.\n4. In the Model field, select a model from the list, or choose Auto to let the system select the default model.\n![Agent card showing the Agent model section with the Model field set to Auto.]\nThe Agent Models list page shows all models available in the environment. Each entry displays the model name, its availability status, and its retirement date (if scheduled). Use this page to understand which models are currently available before making a selection.\n![Agent Models list page showing available models with their status and retirement dates.]\n> TIP\n> If you select Auto, the agent uses whatever model Microsoft designates as the current default. This means the model can change over time without manual intervention. If you need predictable behavior, pin your agent to a specific model version.\n## Model resolution\nWhen an agent runs a task, Business Central resolves which model to use in the following order:\n1. If a model is specified on the task, that model is used.\n2. If no model is set on the task but a model is specified on the agent, the agent-level model is used.\n3. If no model is set on either the task or the agent, the default model is used.\nThis resolution order lets you set a general model preference at the agent level while overriding it for specific tasks that benefit from a different model.\n## Model lifecycle\nModels follow a lifecycle similar to code deprecation in Business Central:\nPreview -> Available -> Obsolete -> Retired\nThe following points apply:\n Multiple models can coexist in the same environment.\n The model lifecycle doesn't align with the standard Business Central deprecation timelines.\n The default model (used by auto mode) can change over time.\n Agents that use auto selection might experience behavior changes when the default model changes.\n Model updates and the availability of new model-selection features in the UI can roll out independently. During transitions, agents continue to work as expected with the updated model, even before the selection UI reflects the change.\n> IMPORTANT\n> When the default model changes, we recommend that you re-evaluate your agents, because agent behavior and accuracy can be affected by the model update.\n## Retired and obsolete models\nWhen a model becomes obsolete or retired, the following indicators help you detect and respond to the change:\n UI indicators\n  Annotations for retired or obsolete models are shown on the agent card and model list.\n Developer communication\n  Model lifecycle changes are communicated through AL Home notifications, documentation updates, and telemetry signals.\n## Related information\nConfigure agents programmatically (preview)<ai-agent-sdk-configuration>\nCoding agents in AL (preview)<ai-agent-sdk-overview>\nDefine and register an agent programmatically (preview)<ai-agent-sdk-define-register>\nOverview (preview)<https://learn.microsoft.com/dynamics365/business-central/dev-itpro/ai/ai-development-toolkit-landing-page>"}