Get Started with Azure AI Services | Open AI and Deployment Models
Table of Contents
- Overview - Azure AI Services
- Kind of Azure AI Services
- Responsible AI Services
- Limited Access Features
- Cognitive Account – Open AI
- IaC Deployment – Terraform
- Cognitive Account Purge
- Important Links
Overview - Azure AI Services
- Azure AI services help developers and organizations rapidly create intelligent, cutting-edge, market-ready, and responsible applications with out-of-the-box and prebuilt and customizable APIs and models.
- Azure AI services and Azure Machine Learning both have the end-goal of applying artificial intelligence (AI) to enhance business operations, though how each provides this in the respective offerings is different. Generally, the audiences are different:
- Azure AI services are for developers without machine-learning experience.
- Azure Machine Learning is tailored for data scientists.
- Azure AI services are earlier termed as Cognitive Services but now Cognitive Services and Applied AI Services are coined as Azure AI Services.
- Example applications include natural language processing for conversations, search, monitoring, translation, speech, vision, and decision-making.
Kinds of Azure AI Services
- Many of the Azure AI services have a free tier you can use to try the service. To use the free tier, use F0 as the SKU for your resource.
Service Description Anomaly Detector (retired) Identify potential problems early on. Azure AI Search Bring AI-powered cloud search to your mobile & web apps. Azure OpenAI Perform a wide variety of natural language tasks. Bot Service Create bots and connect them across channels. Content Moderator (retired) Detect potentially offensive or unwanted content. Content Safety An AI service that detects unwanted contents. Custom Vision Customize image recognition for your business. Document Intelligence Turn documents into intelligent data-driven solutions. Face Detect and identify people and emotions in images. Immersive Reader Help users read and comprehend text. Language Build apps with industry-leading natural language understanding capabilities. Language understanding (retired) Understand natural language in your apps. Metrics Advisor (retired) An AI service that detects unwanted contents. Personalizer (retired) Create rich, personalized experiences for each user. QnA maker (retired) Distill information into easy-to-navigate questions and answers. Speech Speech to text, text to speech, translation, and speaker recognition. Translator Use AI-powered translation technology to translate more than 100 in-use, at-risk, and endangered languages and dialects. Video Indexer Extract actionable insights from your videos. Vision Analyze content in images and videos.
- With Azure and Azure AI services, you have access to a broad ecosystem, such as:
- Automation and integration tools like Logic Apps and Power Automate.
- Deployment options such as Azure Functions and the App Service.
- Azure AI services Docker containers for secure access.
- Tools like Apache Spark, Azure Databricks, Azure Synapse Analytics, and Azure Kubernetes Service for big data scenarios.
Responsible AI Services
- Responsible Artificial Intelligence (Responsible AI) is an approach to developing, assessing, and deploying AI systems in a safe, trustworthy, and ethical way.
- Responsible AI can help proactively guide several decisions toward more beneficial and equitable outcomes, keeping people and their goals at the center of system design decisions and respecting enduring values like fairness, reliability, and transparency.
- Responsible AI Practices - Empowering responsible AI practices | Microsoft AI
Responsible AI Services within Azure AI Services suite
Vision
- Azure AI Vision - Image Analysis
- Azure AI Vision - OCR
- Azure AI Vision - Face
- Azure AI Vision - Spatial Analysis
- Azure Custom Vision
- Azure Video Indexer
Language
- Azure AI Language
- Azure AI Language - Custom text classification
- Azure AI Language - Named entity recognition
- Azure AI Language - Custom named entity recognition
- Azure AI Language - Entity linking
- Azure AI Language - Language detection
- Azure AI Language - Key phrase extraction
- Azure AI Language - Personally identifiable information detection
- Azure AI Language - Question Answering
- Azure AI Language - Sentiment Analysis and opinion mining
- Azure AI Language - Text Analytics for health
- Azure AI Language - Summarization
- Language Understanding
Speech
- Azure AI Speech - Pronunciation Assessment
- Azure AI Speech - Speaker Recognition
- Azure AI Speech - Text to speech
- Azure AI Speech - Speech to text
Search
Other
- Azure OpenAI
- Azure AI Content Safety
- Azure AI Document Intelligence
- Anomaly Detector
- Personalizer
- QnA Maker
Responsible AI terms acceptance
- Why accept Responsible AI terms? - AI systems involve technology, users, impacts, and deployment context. Crafting effective systems requires understanding technology, capabilities, and context. Microsoft's Transparency Notes clarify our AI's workings, choices, and holistic perspective. They aid in system development, deployment, and communication.
- In addition to the Transparency Note, Microsoft offer guidance and resources for responsibly utilizing Azure OpenAI models, aligning with the Microsoft Responsible AI Standard followed by the engineering teams.
- Roles to Accept terms - Azure account must have a Cognitive Services Contributor role assigned in order for you to agree to the responsible AI terms and create a resource.
- How to Execute? - If you're planning to use Spatial Analysis in Azure AI Vision or Text Analytics for Health in Azure AI Language, then you must create your first Vision or Language resources from the Azure portal so you can review and acknowledge the terms and conditions. Below are few examples of that.
- E.g. Computer Vision
- E.g. Azure AI Language
Limited Access Features for Azure AI Services
- Why Limited Access? - Microsoft vision is to empower developers and organizations to use AI to transform society in positive ways. We encourage responsible AI practices to protect the rights and safety of individuals. To achieve this, Microsoft has implemented a Limited Access policy grounded in our AI Principles to support responsible deployment of Azure services.
- How to Access Limited Features? - Limited Access services require registration, and only customers managed by Microsoft—meaning those who are working directly with Microsoft account teams—are eligible for access.
- The use of these services is limited to the use case selected at the time of registration. Customers must acknowledge that they've reviewed and agree to the terms of service. Microsoft may require customers to reverify this information.
- The following Azure AI services have Limited Access:
- Custom Neural Voice: Pro features
- Custom text to speech avatar: All features
- Speaker Recognition: All features
- Face API: Identify and Verify features
- Azure AI Vision: Celebrity Recognition feature
- Azure AI Video Indexer: Celebrity Recognition and Face Identify features
- Azure OpenAI: Azure OpenAI Service, modified abuse monitoring, and modified content filters
Azure Cognitive Account - Kind Open AI
- Open AI Services
- Azure OpenAI Service provides REST API access to OpenAI's powerful language models including the GPT-3.5-Turbo model series, GPT-4, GPT-4 Turbo with Vision, GPT-4o & GPT-4 Turbo NEW and Embeddings model series.
- These models can be easily adapted to your specific task including but not limited to content generation, summarization, image understanding, semantic search, and natural language to code translation.
- Users can access the service through REST APIs, Python SDK, or our web-based interface in the Azure OpenAI Studio.
-
How to Access OPEN AI - Access is currently limited due to high demand, upcoming product improvements, and Microsoft’s commitment to responsible AI. Azure OpenAI requires registration and is currently only available to approved enterprise customers and partners. You can apply here for access: Apply now
- Open AI Deployment Models
- Once you create an Azure OpenAI Resource, you must deploy a model before you can start making API calls and generating text.
- This action can be done using the Deployment APIs. These APIs allow you to specify the model you wish to use.
-
Azure OpenAI Service is powered by a diverse set of models with different capabilities and price points. Model availability varies by region
Models Description GPT-4o & GPT-4 Turbo NEW The latest most capable Azure OpenAI models with multimodal versions, which can accept both text and images as input. GPT-4 A set of models that improve on GPT-3.5 and can understand and generate natural language and code. GPT-3.5 A set of models that improve on GPT-3 and can understand and generate natural language and code. Embeddings A set of models that can convert text into numerical vector form to facilitate text similarity. DALL-E A series of models that can generate original images from natural language. Whisper A series of models in preview that can transcribe and translate speech to text. Text to speech (Preview) A series of models in preview that can synthesize text to speech. - The default quota for models varies by model and region. Default quota limits are subject to change. Select the appropriate model along with the correct version to deploy OPEN AI Models.
-
Model summary table and region availability - Azure OpenAI Service models - Azure OpenAI | Microsoft Learn
- Azure OpenAI now supports automatic updates for select model deployments.
- Roles required: Roles and permissions
- Cognitive Services Open AI Contributor role required to edit model and deployments
- Cognitive Services Usages Reader role required for Accessing quota for model deployments.
- Sample Screenshot of Open AI Deployment Models
- Open AI Studio
- Azure AI Studio is a trusted and inclusive platform that empowers developers of all abilities and preferences to innovate with AI and shape the future. You can play with configured deployment models in Azure Open AI Studio and update scale, capacity settings for different models.
- With Azure AI Studio, you can evaluate large language model (LLM) responses and orchestrate prompt application components with prompt flow for better performance.
- The platform facilitates scalability for transforming proof of concepts into full-fledged production with ease like you can build generative AI applications on an enterprise-grade platform and seamlessly explore, build, test, and deploy using cutting-edge AI tools and ML models, grounded in responsible AI practices.
- Sample Screenshot of Azure Open AI Studio
IaC Deployment (Terraform)
- Cognitive Account Implementation
sku_name
varies based on different kinds of Azure AI Services- Possible values of Kind are:
Academic
,AnomalyDetector
,Bing.Autosuggest
,Bing.Autosuggest.v7
,Bing.CustomSearch
,Bing.Search
,Bing.Search.v7
,Bing.Speech
,Bing.SpellCheck
,Bing.SpellCheck.v7
,CognitiveServices
,ComputerVision
,ContentModerator
,ContentSafety
,CustomSpeech
,CustomVision.Prediction
,CustomVision.Training
,Emotion
,Face
,FormRecognizer
,ImmersiveReader
,LUIS
,LUIS.Authoring
,MetricsAdvisor
,OpenAI
,Personalizer
,QnAMaker
,Recommendations
,SpeakerRecognition
,Speech
,SpeechServices
,SpeechTranslation
,TextAnalytics
,TextTranslation
andWebLM
. - Few IaC kinds mapped to Azure AI Services:
S.N. Azure AI Services Kind 1 Open AI (Limited Access) OpenAI 2 Azure AI Content Safety ContentSafety 3 Azure AI Translation TextTranslation 4 Azure AI Speech SpeechServices 5 Azure AI Vision (Responsible AI Service) ComputerVision 6 Azure AI language (Responsible AI Service) TextAnalytics 7 Azure AI Document Intelligence FormRecognizer 8 Azure AI services multi-service account CognitiveServices - Sample Terraform code shared below with Kind as OpenAI
resource "azurerm_cognitive_account" "example" { name = "example-account" location = azurerm_resource_group.example.location resource_group_name = azurerm_resource_group.example.name kind = "OpenAI" sku_name = "S0" tags = { Acceptance = "Test" } }
- Cognitive Model Deployment Implementation
- Model Deployment is only supported for Open AI format
cognitive_account_id
is required parameter where you can pass ID of the above created Cognitive Account- Pass the values of parameters like
model
(Required),scale
(Required) - Refer list of models and its version available in a region - Azure OpenAI Service models - Azure OpenAI | Microsoft Learn and select the right configuration
- Sample Terraform code shared below with format as OpenAI
resource "azurerm_cognitive_deployment" "example" { name = "example-cd" cognitive_account_id = azurerm_cognitive_account.example.id model { format = "OpenAI" name = "text-curie-001" version = "1" } scale { type = "Standard" } }
Cognitive Account Purge
- Once you delete a Cognitive Account, you won't be able to create another one with the same name for 48 hours. To create a resource with the same name, you need to purge the deleted resource.
- You must have necessary permissions to purge a resource. The least permission a subscription must have -
Microsoft.CognitiveServices/locations/resourceGroups/deletedAccounts/delete
to purge resources, such as Cognitive Services Contributor or Contributor. - When using Contributor to purge a resource the role must be assigned at the subscription level. If the role assignment is only present at the resource or resource group level, you can't access the purge functionality.
- Recover or purge deleted Azure AI services resources - Azure AI services | Microsoft Learn
Important Links
Overview - Azure AI Services
- Azure AI services
- Azure OpenAI Service
- Transparency note for Spatial Analysis
- Create a multi-service resource for Azure AI services
- Azure AI services documentation
Responsible AI
- Overview of Responsible AI
- Responsible use of AI with Azure AI services
- Empowering responsible AI practices
Limited Access Features
Azure Open AI Service and Deployment Model
- Azure OpenAI Service
- Overview of Responsible AI practices for Azure OpenAI models
- Azure OpenAI Service models
- Working with Azure OpenAI models
- Standard deployment model availability
- Azure AI Studio
- Role-based access control for Azure OpenAI Service
Cognitive Account - IaC
Published on:
Learn moreRelated posts
How Azure AI Search powers RAG in ChatGPT and global scale apps
Millions of people use Azure AI Search every day without knowing it. You can enable your apps with the same search that enables retrieval-augm...
Episode 388 – Getting Started with Azure Bicep: Infrastructure as Code with a Domain Specific Language
Welcome to Episode 388 of the Microsoft Cloud IT Pro Podcast. In this episode, we dive into Azure Bicep, Microsoft’s streamlined language for ...
RAG with SQL Vector Store: A Low-Code/No-Code Approach using Azure Logic Apps
Data is at the heart of every AI application, and efficient data ingestion is critical for success. With over 1,400 enterprise connectors, Log...
Exciting Announcement: Public Preview of Native Vector Support in Azure SQL Database!
Public Preview of Native Vector Support in Azure SQL Database We are excited to share that the dedicated vector data type in Azure SQL Databas...
Scaling Azure Container Apps for Peak Performance
Unlock the full potential of Azure Container Apps with smart scaling! In this final part of our series, explore how to keep your apps responsi...