Model router
Model router
Version: 2025-08-07
MicrosoftLast updated August 2025
Model router is a deployable AI model that is trained to select the most suitable large language model (LLM) for a given prompt.
Multipurpose
Multimodal
Welcome to the 2025-08-07 version of model router that supports gpt-5 models in addition to the previously supported gpt-4 and o4-mini models. This version is currently limited access only. If you already have o3 or gpt-5 access, no request is required.
Otherwise, you can register for access using the gpt-5 and model router limited access application form .
Note: If you are not yet registered, the previous version, 2025-05-19 with support for the gpt-4.1 and o4-mini models will be offered for deployment.
Model Router dynamically selects the optimal large language model (LLM) for a specific query or task in real time. By evaluating factors like query complexity, cost, and performance, it efficiently routes requests to the most suitable model, ensuring high quality results while minimizing costs. Context length for model router is dependent on the underlying model that's being used for each prompt. For more information, reference the Model router documentation

Model version in the router

Router VersionModelModel versionAvailabilityLifecycle
2025-08-07gpt-52025-08-07Global standardGeneral available
gpt-5-mini2025-08-07Global standardGeneral available
gpt-5-nano2025-08-07Global standardGeneral available
gpt-5-chat2025-08-07Global standardPreview
gpt-4.12025-04-14Global standardGeneral available
gpt-4.1-mini2025-04-14Global standardGeneral available
gpt-4.1-nano2025-04-14Global standardGeneral available
o4-mini2025-04-16Global standardGeneral available
Model Specifications
Context Length1048576
LicenseCustom
Training DataAugust 2025
Last UpdatedAugust 2025
Input TypeText,Image
Output TypeText
PublisherMicrosoft
Languages1 Language