Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
On November 2, 2021, Azure Language was released into public preview. Azure Language unifies the Text Analytics, QnA Maker, and Language Understanding (LUIS) service offerings, and provides several new features as well.
Important
- As of March 31, 2026, LUIS runtime and authoring endpoints (including REST API calls) are fully retired, and all LUIS requests fail.
- To ensure uninterrupted operation, export all of your LUIS app data as soon as possible in order to:
- Start using Conversational Language Understanding (CLU), or
- Try out a Language Studio.
- Use the following API to export your data: LUIS Versions - Export.
How do I migrate to Azure Language if I'm using Language Understanding (LUIS)?
If you're using Language Understanding (LUIS), you can import your Language Understanding (LUIS) JSON file to the new Conversational language understanding feature:
Export the LUIS app
- Export your LUIS application as JSON by using the LUIS authoring APIs or the LUIS portal.
- LUIS Versions - Export REST API
Map LUIS data to the CLU schema
- Update the exported JSON to align intents, entities, and utterances with the CLU project schema.
- Conversational Language Understanding overview provides details on the CLU schema and supported features.
Create a CLU project and import data
- Create a CLU project and import the prepared data by using Language Studio or the CLU authoring APIs.
- CLU quickstart (REST API) provides instructions on creating a CLU project and importing data.
Train the model
- Start a training job to build the CLU model from the imported data.
- Train and evaluate a CLU model provides details on training and evaluating your CLU model.
Test the model
- Send test utterances to the CLU prediction endpoint and verify intent and entity results.
- Conversational Language Understanding overview provides details on querying your CLU model.