The Dimensions Semantic Search MCP server gives AI assistants access to annotated scientific document repositories - Dimensions publications, grants, patents, and clinical trials. Documents are enriched with ontological concept annotations across 40+ life science domains, enabling searches that automatically resolve terms to all related subtypes and compounds. Querying “cancer“ finds every indexed subtype; querying “PFAS“ resolves to PFOS, PFOA, PFHxS, and all related compounds automatically. Currently supported by all major AI platforms like Claude Desktop, VS Code, Cursor, ChatGPT and many more.
Example Prompts
Once configured, try asking your AI assistant:
Data Schema
Describe the dspub repository schema
Which vocabularies / concept domains are available, e.g. diseases, drugs, proteins?
Which synonyms are known for aspirin?
Document Search
Search Dimensions Publications for recent papers about PFAS compounds and hepatotoxicity
Find patents mentioning aspirin and headache within 30 words of each other. Sort by relevance.
Analytics
Which drugs most frequently co-occur with Alzheimer's disease?
How many documents mention both CRISPR and sickle cell disease published after 2018?
How often does psoriasis appear in https://doi.org/10.1038/s41435-026-00402-5?
Search Including Metadata
For which clinical phase and corresponding disease are tyrosine kinase inhibitors investigated in running clinical trials?
Which drugs for Breast Cancer treatment were mentioned in publications in the last 6 months?
What are the key opinion leaders on the topic of Gut-Muscle Axis?
What are the most referenced authors and their affiliated organizations in psoriasis-related publications of the past 3 months?
What are the sponsors for studies about afatinib treatment?
I want to find relevant publications for anthraquinones as potential active pharmaceutical ingredients, filtered by the Field of Research metadata field for “Pharmacology and Pharmaceutical Sciences“. Look up the correct code before searching.
What critical research funding can be observed for which funding organizations in the research area of aging and longevity?
Available Tools
The Dimensions Semantic Search MCP exposes tools designed to be called in natural sequence: tools for discovery (listing available repositories, concept domains, and document schemas); concept resolution (resolving terms and retrieving ontological relationships); search and retrieval (structured concept and full-text queries with filtering, pagination, and sorting); and analysis (co-occurrence queries across the pre-computed sentence-level index).
View the full tool referenceSetup
This MCP connects via HTTP - no local binary or npm package required.
- Get your API key - If your organization has a subscription, you can retrieve your API key here.
- Install the MCP server - Configuration instructions are available here.
- Restart your client - Restart your AI client to activate the server.
Limitations
The Dimensions Semantic Search MCP is built on the Digital Science Knowledge Extractor Semantic Search API, which defines the scope of available repositories and query structures. Available document repositories and supported features vary - use the discovery tools to confirm what is accessible with your credentials. Technical rate limits apply to all requests to ensure platform stability. By using this MCP, users agree to our reasonable use policy.
