Data API
The Moonlit Data API gives applications direct access to the same data layer that powers the MCP server. Search, retrieve, use the reference graph, and stream AI research from your own backend.
How it differs from MCP
The Data API and the MCP server are the two access modes within Moonlit Connect. Both read from the same data layer.
Use the MCP server when you want Claude or any MCP-compatible client to research European law. Use the Data API when you are integrating the data layer into your own backend, building a custom RAG pipeline, or operating an LLM-agnostic product surface. Access to the Data API is set up by our team; see Get access.
Capabilities
Six search modes: keyword (with Boolean operators and sort options), semantic, hybrid, semantic-reranked, hybrid-reranked, and reference search. Five filter endpoints with hierarchical taxonomies for jurisdictions, document types, fields of law, sources, and semantic-enabled portals. Full document retrieval with structured metadata, citation counts, in-force status, and AI-generated summaries. Article-level traversal for composite documents.
Luna
Luna is a streaming legal research assistant exposed as a Data API endpoint pair: initialize a session with initialize_chat, then send a question with optional jurisdiction scoping to ask_question and receive the answer with inline citations over Server-Sent Events. Every citation resolves to a document in the data layer.
Get started
- Quickstart: your first request in three steps.
- Reference: every endpoint, parameter, and response field, generated from the live OpenAPI spec.
- Integration patterns: common shapes for monitoring, compliance research, and reference graph analytics.
Endpoint base
All Data API endpoints sit under https://api.moonlit.ai/v1.1, authenticated with the Ocp-Apim-Subscription-Key header. The reference documents each endpoint with cURL, Python, and JavaScript examples generated from the live OpenAPI spec.