Azure Language in Foundry Tools
Build apps that understand, analyze, and generate human language with AI.
Overview
Add advanced natural language processing to your generative AI apps
- Extract personal data or entities, summarize conversations or articles, identify language and sentiment, and analyze health records across text and documents using state-of-the-art transformer models.
- Build small or large language models to analyze text, identify intents, answer questions and extract more entities for your own domain. Train the AI model with one language, then use it for multiple other languages.
- Run AI models wherever your data resides and deploy your apps in the cloud or at the edge with containers. Build multilingual assistants and chatbots with generative AI models and Azure Translator.
- Build higher quality generative AI solutions with lower latency using Azure Language and Azure OpenAI together. Protect privacy with personal data detection. Segment and summarize long conversations to improve context management for your agents. Reduce inaccuracies with named entity recognition and orchestrate conversational apps more efficiently with conversational language understanding (CLU).
FEATURES
Augment your generative AI app with best-in-class language models
Protect personal sensitive data
Detect and redact personal data before storing or retrieving it in your application workflows.
Extract named entities
Recognize named entities, use predefined entities, or customise your own. Extract the exact text values and avoid introducing inaccuracies.
Summarize content
Derive summaries from documents, meetings, and other sources. Segment long conversations based on topics and extract key sentences for more accurate summarizations.
Customise your conversational AI language models
Build your agents with CLU and the ability to answer custom questions. Customise multilingual models that identify intentions and extract information for efficient agent routing—and create bots to answer questions about your business.
Analyze health data
Convert unstructured text such as doctor's notes, discharge summaries, clinical documents, and electronic health records into actionable insights.
Security
Embedded security and compliance
34,000
Full-time equivalent engineers dedicated to security initiatives at Microsoft.
15,000
Partners with specialized security expertise.
>100
Compliance certifications, including over 50 specific to global regions and countries.
Pricing
Flexible pricing to meet your needs
Pay for only what you use—no upfront costs. Pay-as-you-go pricing is based on:
CUSTOMER STORIES
Trusted by companies of all sizes
RESOURCES
Get started with Azure Language
Azure Language documentation
Explore tutorials, API references, and other Language resources.
Azure Language developer guide
Explore sample code, SDKs, and REST APIs for Language.
Try it out in Azure AI Foundry
Discover the prebuilt capabilities or customise models quickly in Azure AI Foundry.
FAQ
Frequently asked questions
-
Azure Language (formerly Azure AI Language) offers advanced natural language processing (NLP) capabilities out of the box and customizable AI models including personal data redaction, entity extraction, summarization, intent classification, question answering, text analytics for health, etc. Azure Language uses state-of-the-art transformer models with large and small language models (i.e. LLMs and SLMs), making it easier to develop intelligent multilingual applications through scalable and reliable APIs, SDK, and Azure AI Foundry.
-
Yes, Azure AI Language—formerly part of Azure AI Services—is now rebranded as Azure Language in Foundry Tools. This change is part of a broader platform unification under Azure AI Foundry, reflecting how developers increasingly use these services as modular tools to build intelligent, agentic applications.
Azure Language still offers the same powerful capabilities—like conversational language understanding (CLU), custom question answering, summarization, named entity recognition, and health data analysis. What’s new is its positioning: Language is now a core tool within a unified agent-building Foundry platform and offers developers a choice of large language models.
This shift makes it easier for developers to discover, orchestrate, and integrate natural language capabilities into modern AI workflows—whether you're building multilingual chatbots, analyzing documents, or powering intelligent assistants. -
Azure AI Language empowers developers to build language AI apps at scale with high quality and low latency. It reduces prompt engineering through prebuilt capabilities on fine-tuned models. PII redaction protects privacy data. Entity extraction enables factual checks to reduce inaccuracies. Summarization offers options to chapter meeting per topic. Conversational language understanding and custom question answering make it easy to build bots that classify user intents and answer questions.
-
Try out the prebuilt capabilities in Azure AI Foundry, without requiring an account. Then create your Azure AI resource in the studio to access all the capabilities to build your apps. Check out Azure Language documentation for more details about each capability, and the developer guide to get access to client libraries, REST API and sample code.
-
See each feature's supported languages in the Azure Language documentation for more information.
-
Yes, use the analyze operation to combine more than one feature in the same asynchronous call. The analyze operation is currently only available in the Standard pricing tier and follows the same pricing criteria.
-
As of October 31, 2025, the QnA Maker service was fully retired. This means:
All runtime and inferencing requests will fail. Any applications or bots relying on QnA Maker will stop functioning. QnA Maker projects, including knowledge bases, can no longer be accessed or updated. Users should transition to Custom Question Answering (CQA) in Azure Language to maintain continuity. Migration tools and documentation are available to help move existing QnA Maker content to supported services. It’s strongly recommended to do so before the retirement date to avoid service disruption. - Yes, Microsoft supports custom intent classification using Conversational Language Understanding (CLU). Customers are migrating from Language Understanding Intelligence Service (LUIS) to CLU to take advantage of the high-quality, deterministic intent classification in CLU.
Microsoft is retiring LUIS in phases:
LUIS portal access ends October 31, 2025. After this date, you can no longer use the portal to manage apps. Only REST API authoring and inferencing will remain available. Full retirement occurs March 31, 2026. After this date, all LUIS services, including authoring and inferencing APIs, will be shut down.
Next steps
Choose the Azure account that’s right for you
Pay as you go or try Azure free for up to 30 days.
AI development tools
Design and manage AI applications
Create, customise, and scale AI apps and agents efficiently.
Business Solutions Hub
Drive results with innovative cloud solutions
Browse the Business Solutions Hub to find products and solutions to achieve your goals.
Get the Azure mobile app 
