microsoft/Recognizers-Text

Health v2 · maintenance only

Adoption not part of the score

1794 stars · 434 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Microsoft.Recognizers-Text is a multilingual library for recognizing and resolving entities such as numbers, units, and date/time expressions in text. It is available as packages for .NET, JavaScript/TypeScript, Python, and Java, and powers pre-built entities in LUIS, Power Virtual Agents, and the Microsoft Bot Framework.

Use cases

• extract dates and times from user text in a chatbot
• parse numbers and units from natural language input
• recognize entities in multiple languages like English, Chinese, and Spanish
• normalize datetime expressions to structured values
• add pre-built entity recognition to a LUIS or Bot Framework app
• extract number expressions from free-form text

When to choose

• you need robust, multilingual extraction of numbers, units, or date/time from text
• you are building a bot or conversational app on Microsoft Bot Framework or LUIS
• you want a deterministic parser rather than an ML model for entity extraction
• you need packages across .NET, JavaScript, Python, or Java

When to avoid

• you need general-purpose NER for people, organizations, or locations
• you need an ML-based or deep-learning entity extractor
• your target language is only partially supported (e.g., Japanese, Korean, Arabic, Swedish) and support quality matters
• you need a single-language-only lightweight solution

Facets

library · maturity stable

nlp parser sdk developer-tools apis dotnet python jvm cross-platform entity-extraction named-entity-recognition datetime-recognition number-recognition multilingual bot-framework natural-language-processing nodejs

Member repositories

For agents

markdown · JSON · MCP: product_card(name="microsoft/Recognizers-Text")