Azure AI Language
Create high-quality copilots and generative AI applications for text, documents, and conversations using task-optimized, prebuilt, customizable small and large language models.
Overview
Add 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 Microsoft Azure AI Translator.
- Build higher quality generative AI solutions with lower latency using Azure AI Language and Azure OpenAI together. Protect privacy with personal data detection. 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-of-breed 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 customize 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.
Customize your conversational AI language models
Build your copilot apps with CLU and custom question answering to customize 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
Built-in security and compliance
Microsoft has committed to investing $20 billion in cybersecurity over five years.
We employ more than 8,500 security and threat intelligence experts across 77 countries.
Azure has one of the largest compliance certification portfolios in the industry.
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 AI Language
Azure AI Language documentation
Explore tutorials, API references, and other Azure AI Language resources.
Azure AI Language developer guide
Explore sample code, SDKs, and REST APIs for Azure AI Language.
Try it out in Azure AI Studio
Discover the prebuilt capabilities or customize models quickly in Azure AI Studio.
FAQ
Frequently asked questions
-
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. The service 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 Studio.
-
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 Studio, 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 AI 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 AI 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.