Semantic Scholar MCP Server

Semantic Scholar MCP server enabling AI agents to search and retrieve academic papers from Semantic Scholar — Allen Institute for AI's free academic search engine covering 200M+ papers across all fields. Enables searching by keyword, author, or paper ID, retrieving paper metadata, abstracts, citation counts, and open-access PDFs, integrating academic literature search into research agent workflows.

Evaluated Mar 06, 2026 (0d ago) vcurrent
Homepage ↗ Repo ↗ Other semantic-scholar academic research papers mcp-server citations ai2 science
⚙ Agent Friendliness
75
/ 100
Can an agent use this?
🔒 Security
82
/ 100
Is it safe for agents?
⚡ Reliability
72
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
70
Documentation
72
Error Messages
70
Auth Simplicity
90
Rate Limits
82

🔒 Security

TLS Enforcement
95
Auth Strength
80
Scope Granularity
72
Dep. Hygiene
75
Secret Handling
85

HTTPS. Optional API key. Academic metadata is public. No sensitive data. Free AI2 API.

⚡ Reliability

Uptime/SLA
75
Version Stability
72
Breaking Changes
70
Error Recovery
70
AF Security Reliability

Best When

A researcher or agent needs to search scientific literature — Semantic Scholar's free API provides 200M+ papers with semantic search, citations, and open-access PDFs without the copyright concerns of Sci-Hub.

Avoid When

You need full-text access to paywalled papers — use your institution's library database for legitimate full-text access.

Use Cases

  • Searching academic literature for research topics from literature review agents
  • Retrieving paper abstracts and metadata for citation analysis from research agents
  • Finding related papers and citation networks from knowledge discovery agents
  • Accessing open-access full-text papers for analysis from reading agents
  • Identifying top researchers and institutions in a field from research intelligence agents
  • Tracking citation counts and paper influence from bibliometrics agents

Not For

  • Accessing paywalled papers (Semantic Scholar only surfaces metadata/open access; use institutional library access)
  • Real-time preprint monitoring (use arXiv MCP for very recent preprints)
  • Patents and technical standards (Semantic Scholar focuses on peer-reviewed literature)

Interface

REST API
Yes
GraphQL
No
gRPC
No
MCP Server
Yes
SDK
No
Webhooks
No

Authentication

Methods: api_key none
OAuth: No Scopes: No

Semantic Scholar API: free without API key (limited rate); API key for higher limits. Register at semanticscholar.org/product/api for free API key. Set S2_API_KEY environment variable.

Pricing

Model: free
Free tier: Yes
Requires CC: No

Semantic Scholar API is completely free (Allen Institute for AI). High-volume partners can request higher limits. MCP server is free.

Agent Metadata

Pagination
offset
Idempotent
Full
Retry Guidance
Not documented

Known Gotchas

  • Without API key, rate limits are strict (5000 req/5min) — register for free key for agent use
  • Paper coverage varies by field — CS/AI papers very comprehensive; some humanities fields less so
  • Open access PDFs available for ~50% of papers — others have metadata only
  • Semantic Scholar IDs differ from DOIs — translate between ID systems as needed
  • Author disambiguation is imperfect — verify author identity for important citation queries
  • Search relevance is good but semantic search quality varies by field and query complexity

Alternatives

Full Evaluation Report

Detailed scoring breakdown, competitive positioning, security analysis, and improvement recommendations for Semantic Scholar MCP Server.

$99

Scores are editorial opinions as of 2026-03-06.

5178
Packages Evaluated
26151
Need Evaluation
173
Need Re-evaluation
Community Powered