Merge pull request #4268 from MicrosoftDocs/main

4/22/2025 PM Publish

Author: Taojunshen

articles/ai-foundry/model-inference/overview.md

@@ -18,30 +18,27 @@ recommendations: false
 
 Azure AI model inference provides access to the most powerful models available in the Azure AI model catalog. The models come from key model providers in the industry, including OpenAI, Microsoft, Meta, Mistral, Cohere, G42, and AI21 Labs. These models can be integrated with software solutions to deliver a wide range of tasks that include content generation, summarization, image understanding, semantic search, and code generation.
 
-> [!TIP]
-> To deploy DeepSeek-R1 or OpenAI o3-mini in Azure AI model inference, follow the steps at [Add and configure models](how-to/create-model-deployments.md).
-
 Azure AI model inference provides a way to **consume models as APIs without hosting them on your infrastructure**. Models are hosted in a Microsoft-managed infrastructure, which enables API-based access to the model provider's model. API-based access can dramatically reduce the cost of accessing a model and simplify the provisioning experience.
 
 Azure AI model inference is part of Azure AI Services, and users can access the service through [REST APIs](./reference/reference-model-inference-api.md), [SDKs in several languages](supported-languages.md) such as Python, C#, JavaScript, and Java. You can also use the Azure AI model inference from [Azure AI Foundry by configuring a connection](how-to/configure-project-connection.md).
 
 ## Models
 
-You can get access to the key model providers in the industry including OpenAI, Microsoft, Meta, Mistral, Cohere, G42, and AI21 Labs. Model providers define the license terms and set the price for use of their models. The following list shows all the models available:
-
-To see details for each model including, language, types, and capabilities, see [Models](concepts/models.md) article.
-
-| Provider | Models |
-| -------- | ------ |
-| [AI21 Labs](concepts/models.md#ai21-labs) | - AI21-Jamba-1.5-Mini <br /> - AI21-Jamba-1.5-Large <br /> |
-| [Azure OpenAI](concepts/models.md#azure-openai) | - o3-mini <br /> - o1 <br /> - gpt-4o <br /> - o1-preview <br /> - o1-mini <br /> - gpt-4o-mini <br /> - text-embedding-3-large <br /> - text-embedding-3-small <br /> |
-| [Cohere](concepts/models.md#cohere) | - Cohere-embed-v3-english <br /> - Cohere-embed-v3-multilingual <br /> - Cohere-command-r-plus-08-2024 <br /> - Cohere-command-r-08-2024 <br /> - Cohere-command-r-plus <br /> - Cohere-command-r <br /> |
-| [Core42](concepts/models.md#core42) | - jais-30b-chat <br /> |
-| [DeepSeek](concepts/models.md#deepseek) | - DeepSeek-V3 <br /> - DeepSeek-R1 <br /> |
-| [Meta](concepts/models.md#meta) | - Llama-3.3-70B-Instruct <br /> - Llama-3.2-11B-Vision-Instruct <br /> - Llama-3.2-90B-Vision-Instruct <br /> - Meta-Llama-3.1-405B-Instruct <br /> - Meta-Llama-3-8B-Instruct <br /> - Meta-Llama-3.1-70B-Instruct <br /> - Meta-Llama-3.1-8B-Instruct <br /> - Meta-Llama-3-70B-Instruct <br /> |
-| [Microsoft](concepts/models.md#microsoft) | - Phi-4-multimodal-instruct <br /> - Phi-4-mini-instruct <br />  - Phi-4 <br /> - Phi-3-mini-128k-instruct <br /> - Phi-3-mini-4k-instruct <br /> - Phi-3-small-8k-instruct <br /> - Phi-3-medium-128k-instruct <br /> - Phi-3-medium-4k-instruct <br /> - Phi-3.5-vision-instruct <br /> - Phi-3.5-MoE-instruct <br /> - Phi-3-small-128k-instruct <br /> - Phi-3.5-mini-instruct <br /> |
-| [Mistral AI](concepts/models.md#mistral-ai) | - Ministral-3B <br /> - Mistral-large <br /> - Mistral-small <br /> - Mistral-Nemo <br /> - Mistral-large-2407 <br /> - Mistral-Large-2411 <br /> - Codestral-2501 <br /> |
-| [NTT Data](concepts/models.md#ntt-data) | - Tsuzumi-7b |
+You can get access to the key model providers in the industry including OpenAI, Microsoft, Meta, Mistral, Cohere, G42, and AI21 Labs. Model providers define the license terms and set the price for use of their models.
+
+Explore the following model families available:
+
+- [AI21 Labs](concepts/models.md#ai21-labs)
+- [Azure OpenAI](concepts/models.md#azure-openai)
+- [Cohere](concepts/models.md#cohere)
+- [Core42](concepts/models.md#core42)
+- [DeepSeek](concepts/models.md#deepseek)
+- [Meta](concepts/models.md#meta)
+- [Microsoft](concepts/models.md#microsoft)
+- [Mistral AI](concepts/models.md#mistral-ai)
+- [NTT Data](concepts/models.md#ntt-data)
+
+To see details for each model including language, types, and capabilities, see [Models](concepts/models.md) article.
 
 ## Pricing
 

articles/ai-services/agents/media/tools/knowledge-tools-list.png

@@ -21,7 +21,7 @@ items:
     href: concepts/jailbreak-detection.md
   - name: Groundedness detection (preview)
     href: concepts/groundedness.md
-  - name: Protected material detection (preview)
+  - name: Protected material detection
     href: concepts/protected-material.md
   - name: Custom categories (preview)
     href: concepts/custom-categories.md

articles/ai-services/content-safety/toc.yml

@@ -103,7 +103,7 @@ These models are currently available for use in Azure OpenAI Service.
 | `gpt-4` | 1106-preview | To be upgraded to **`gpt-4o` version: `2024-11-20`**, starting no sooner than April 17, 2025 **<sup>1</sup>** <br>Retirement date: May 1, 2025  | `gpt-4o`|
 | `gpt-4` | 0125-preview |To be upgraded to **`gpt-4o` version: `2024-11-20`**, starting no sooner than April 17, 2025 **<sup>1</sup>** <br>Retirement date: May 1, 2025  | `gpt-4o` |
 | `gpt-4` | vision-preview | To be upgraded to **`gpt-4o` version: `2024-11-20`**, starting no sooner than April 17, 2025  **<sup>1</sup>** <br>Retirement date: May 1, 2025 | `gpt-4o`|
-| `gpt-4.5-preview` | 2025-02-27 | No earlier than July 02, 2025 | `gpt-4.1` |
+| `gpt-4.5-preview` | 2025-02-27 | July 14, 2025 | `gpt-4.1` |
 | `gpt-4.1` | 2025-04-14 | No earlier than April 11, 2026 | |
 | `gpt-4.1-mini` | 2025-04-14 | No earlier than April 11, 2026 |
 | `gpt-4.1-nano` | 2025-04-14 | No earlier than April 11, 2026 |

articles/ai-services/openai/concepts/model-retirements.md

@@ -41,7 +41,6 @@ At a high level you can break down working with functions into three steps:
 * `gpt-4o-mini` (`2024-07-18`)
 * `gpt-4.5-preview` (`2025-02-27`)
 * `gpt-4.1` (`2025-04-14`)
-* `gpt-4.1-nano` (`2025-04-14`)
 * `gpt-4.1-mini` (`2025-04-14`)
 
 Support for parallel function was first added in API version [`2023-12-01-preview`](https://github.com/Azure/azure-rest-api-specs/blob/main/specification/cognitiveservices/data-plane/AzureOpenAI/inference/preview/2023-12-01-preview/inference.json)
@@ -51,6 +50,7 @@ Support for parallel function was first added in API version [`2023-12-01-previe
 * All the models that support parallel function calling
 * `o4-mini` (`2025-04-16`)
 * `o3` (`2025-04-16`)
+* `gpt-4.1-nano` (`2025-04-14`)
 * `o3-mini` (`2025-01-31`)
 * `o1` (`2024-12-17`)
 * `gpt-4` (`0613`)
@@ -61,6 +61,9 @@ Support for parallel function was first added in API version [`2023-12-01-previe
 > [!NOTE]
 > The `tool_choice` parameter is now supported with `o3-mini` and `o1`. For more information on what parameters are supported with the o-series models see, the [reasoning models guide](./reasoning.md).
 
+> [!IMPORTANT]
+> Tool/function descriptions are currently limited to 1024 characters with Azure OpenAI. We will update this article if this limit is changed.
+
 ## Single tool/function calling example
 
 First we will demonstrate a simple toy function call that can check the time in three hardcoded locations with a single tool/function defined. We have added print statements to help make the code execution easier to follow: