OMOP MCP Server is a Model Context Protocol (MCP) server that maps clinical terminology to Observational Medical Outcomes Partnership (OMOP) concepts using Large Language Models (LLMs). It provides the find_omop_concept tool to map terms to OMOP concepts, validate terminology mappings, search the OMOP vocabulary, and convert between different clinical coding systems. This server supports prompting guidance to improve accuracy, such as specifying the target OMOP field and table name (e.g., measurement_concept_id in the measurement table) and prioritizing vocabularies (e.g., SNOMED preferred or LOINC > SNOMED > RxNorm). The OMOP concept data and fields can be explored through omop_concept_id_fields.json, which lists OMOP fields and tables that store concept IDs. Claude Desktop integration is supported via claude_desktop_config.json, enabling local workflow execution with a configured environment. Setup requires environment variables for API access (Azure OpenAI or OpenAI). The project emphasizes healthcare data standardization and interoperability, returning OMOP metadata such as CONCEPT_ID, CODE, NAME, CLASS, VOCAB, and related URLs for mapped concepts. The repository is Apache 2.0 licensed and designed for researchers and practitioners working with OMOP mappings.