AzureFeeds

New Models Available!  

We’re excited to announce the preview availability of the following Azure OpenAI Service models for use in the Azure AI Agent Service, starting 5/7: 

• o1 

• o3-mini 

• gpt-4.1 

• gpt-4.1-mini 

• gpt-4.1-nano 

Azure OpenAI o-Series Models 

Azure OpenAI o-series models are designed to tackle reasoning and problem-solving tasks with increased focus and capability. These models spend more time processing and understanding the user’s request, making them exceptionally strong in areas like science, coding, and math compared to previous iterations. 

• o1: The most capable model in the o1 series, offering enhanced reasoning abilities. 

• o3 (coming soon): The most capable reasoning model in the o model series, and the first one to offer full tools support for agentic solutions. 

• o3-mini: A faster and more cost-efficient option in the o3 series, ideal for coding tasks requiring speed and lower resource consumption. 

• o4-mini (coming soon): The most efficient reasoning model in the o model series, well suited for agentic solutions. 

Azure OpenAI GPT-4.1 Model Series 

We are excited to share the launch Agents support for the next iteration of the GPT model series with GPT-4.1, 4.1-mini, and 4.1-nano. The GPT-4.1 models bring improved capabilities and significant advancements in coding, instruction following, and long-context processing that is critical for developers.  

What is GPT-4.1?  

GPT-4.1 is the latest iteration of the GPT-4o model, trained to excel at coding and instruction-following tasks. This model will improve the quality of agentic workflows and accelerate the productivity of developers across all scenarios. 

Key features of GPT-4.1 

GPT-4.1 brings several notable improvements:  

• Enhanced coding and instruction following: The model is optimized for better handling of complex technical and coding problems. It generates cleaner, simpler front-end code, accurately identifies necessary changes in existing code, and consistently produces outputs that compile and run successfully.  

• Long context model: GPT-4.1 supports one million token inputs, allowing it to process and understand extensive context in a single interaction. This capability is particularly beneficial for tasks requiring detailed and nuanced understanding as well as multi-step agents that increase context as they operate. 

• Improved instruction following: The model excels at following detailed instructions, especially agents containing multiple requests. It is more intuitive and collaborative, making it easier to work with for various applications. 

Model capabilities 

In addition to the post-training improvements and long context support, GPT-4.1 retains the same API capabilities as the GPT-4o model family, including tool calling and structured outputs.  

Explore GPT-4.1 today  

GPT-4.1 is now available in the AI Foundry Model Catalog, bringing unparalleled advancements in AI capabilities. This release marks a significant leap forward, offering enhanced performance, efficiency, and versatility across a wide array of applications. Whether you’re looking to improve your customer service chatbot, develop cutting-edge data analysis tools, or explore new frontiers in machine learning, GPT-4.1 has something to offer.  

We invite you to delve into the new features and discover how GPT-4.1 can revolutionize your workflows and applications. Explore, deploy, and build applications using these models today in Azure AI Foundry to access this powerful tool and stay ahead in the rapidly evolving world of AI.  

How to use these models in Azure AI Agent Service

Models with Tool-Calling 

To best support agentic scenarios, we recommend using models that support tool-calling. The Azure AI Agent Service currently supports all agent-compatible models from the Azure AI Foundry model catalog. 

To use these models, use the Azure AI Foundry portal to make a model deployment, then reference the deployment name in your agent. For example: 

agent = project_client.agents.create_agent( model=”llama-3″, name=”my-agent”, instructions=”You are a helpful agent”) 

NOTE: This option should only be used for open-source models (e.g., Cepstral, Mistral, Llama) and not for OpenAI models, which are natively supported in the service. This option should also only be used for models that support tool-calling. 

Models without Tool-Calling 

Though tool-calling support is a core capability for agentic scenarios, we now provide the ability to use models that don’t support tool-calling in our API and SDK. This option may be helpful when you have specific use-cases that don’t require tool-calling. 

The following steps will allow you to utilize any chat-completion model that is available through a Serverless API: 

1. Deploy your desired model through Serverless API. Model will show up on your ‘Models + Endpoints’ page. 


2. Click on model name to see model details, where you’ll find your model’s Target URI and Key. 


3. Create a new Serverless connection on ‘Connected Resources’ page, using the Target URI and Key. 


4. Model can now be referenced in your code (Target URI + ‘@’ + Model Name), for example: 

Model=https://Phi-4-mejco.eastus.models.ai.azure.com/@Phi-4-mejco 

Further Exploration 

• Azure AI Foundry Agent Service: Model Support 

• Explore the models in Azure AI Foundry