Connect Semantic Scholar to Gemini Enterprise
Query Semantic Scholar directly from Gemini Enterprise. Arlet links the two through one MCP connection and takes care of API keys.
This server connects your AI client to the Semantic Scholar Academic Graph API and Recommendations API operated by the Allen Institute for AI. It covers 200M+ papers, 80M+ authors and 2B+ citation edges across all disciplines, making it a general-purpose literature tool for market research, due diligence and technology scouting as well as academic work.
Key features:
- search_papers: relevance search with year range, field of study, publication type (Review / MetaAnalysis / ClinicalTrial ...), venue, minimum citation count and open-access filters. Returns citation counts, DOI / arXiv ids, open-access PDF links and a one-sentence TLDR.
- get_paper: full details by Semantic Scholar id, DOI, arXiv id, PubMed id or Corpus id.
- get_paper_citations / get_paper_references: who cites a paper and what it cites, with citation intents (background / method / result), influential flags and citing-context snippets.
- search_authors: researchers by name with affiliations, paper count, total citations and h-index.
- get_recommendations: related papers for a seed paper from Semantic Scholar's recommendation model.
Licensing: data is returned under the Semantic Scholar API License Agreement (ODC-BY 1.0 / CC BY-NC depending on the record) and every response carries the required 'Semantic Scholar' attribution. Abstract text is included only for open-access papers under a CC-BY family license; for other papers the server returns the TLDR and a link instead.
An API key is optional: anonymous access works within a rate limit shared by all unauthenticated users (HTTP 429 when busy), and a free key requested from Semantic Scholar gives you a dedicated allowance.
How to connect Semantic Scholar to Gemini Enterprise
- 1
In the Arlet dashboard, open Data tools, find Semantic Scholar and switch it on. No API key setup is needed.
- 2
Sign in to Gemini Enterprise with your organization Google account and open the gear icon at the bottom left → Connectors → Add MCP Server.
- 3
In the MCP info tab, enter https://mcp.arlet.dev/mcp/ as the Server URL and "Arlet" as the Name. In the Auth tab, enter the OAuth settings from the Arlet docs (Authorization URL / Token URL / Client ID / Client Secret / Scopes) and click Add.
- 4
In the connector list on the chat screen, click Connect on Arlet, sign in with Google and authorize. Arlet can then be enabled with the toggle.
Adding custom connectors (MCP) requires a Gemini Enterprise subscription and is not available in the consumer Gemini app. See the docs for the values to enter in the Auth tab.