OpenAlex MCP Server

An MCP server that connects AI assistants to OpenAlex — the world's largest open catalog of scholarly research. Search 300M+ papers, 90M+ authors, and 109K+ institutions directly from your AI workflow.

What is this?

This is a Model Context Protocol server that bridges your AI assistant to the OpenAlex API. It gives your AI the ability to:

Search and discover scholarly papers with full-text queries and structured filters
Look up any paper by DOI, PMID, or OpenAlex ID
Explore citation networks — what cites what
Find researchers and explore their publication profiles
Browse institutions, journals, and topic taxonomies
Run aggregate statistics on publication trends

Quick Start

Prerequisites

Node.js v18 or later
An MCP client (Claude Desktop, Cursor, Windsurf, Zed, or any MCP-compatible client)

1. Clone or copy

git clone git@github.com:aula-id/mcp-openalex.git
cd mcp-openalex/api-bridge-mcp
npm install

2. Configure (optional but recommended)

Get a free API key at openalex.org/settings/api for higher rate limits.

Set environment variables:

export OPENALEX_API_KEY="your-key-here"
export OPENALEX_MAILTO="your@email.com"  # polite pool — faster responses

Or skip this — the API works without a key, just with lower rate limits.

3. Connect to your MCP client

Add to your MCP client config:

{
  "mcpServers": {
    "openalex": {
      "command": "node",
      "args": ["/path/to/mcp-openalex/api-bridge-mcp/server.js"],
      "env": {
        "OPENALEX_API_KEY": "your-key-here",
        "OPENALEX_MAILTO": "your@email.com"
      }
    }
  }
}

That's it. Your AI assistant can now search scholarly literature.

Available Tools

Search & Discovery

Tool	Description	Credits
`search_works`	Search papers by keyword with filters (year, type, OA status, citations, author, institution, topic)	10
`search_authors`	Find researchers by name with metrics (h-index, institution, citations)	10
`search_institutions`	Find universities and research organizations	10
`search_topics`	Browse the OpenAlex research topic taxonomy	1

Lookup & Detail

Tool	Description	Credits
`get_work`	Full details of a paper by DOI, PMID, or OpenAlex ID	0
`get_author`	Researcher profile with h-index, citation metrics, and affiliation	0

Citation Analysis

Tool	Description	Credits
`get_work_citations`	Works that cite a given paper	1
`get_work_references`	Works cited BY a given paper (its reference list)	1

Author Publications

Tool	Description	Credits
`get_works_by_author`	List an author's publications with filters	1

Analytics

Tool	Description	Credits
`aggregate_works`	Group-by statistics — publication counts by year, type, OA status, institution, topic, country, or language	1

Configuration

Tool	Description	Credits
`set_config`	Set `api_key` or `mailto` at runtime (persists for the session)	0

Tool Details & Examples

`search_works` — Find papers

The primary search tool. Supports full-text queries with boolean logic.

Parameters:

Param	Type	Required	Description
`query`	string	yes	Search terms. Supports boolean: `("deep learning" AND "medical imaging") NOT survey`
`filters`	object	no	Structured filters (see below)
`sort`	string	no	`relevance_score:desc` (default), `cited_by_count:desc`, `publication_date:desc`
`per_page`	number	no	Results per page (1-50, default 10)
`page`	number	no	Page number (default 1)

Available filters:

Filter	Example	Effect
`from_year`	`2020`	Papers published from 2020 onward
`to_year`	`2024`	Papers published up to 2024
`type`	`"article"`, `"review"`, `"article\|review"`	Filter by work type
`open_access`	`true`	Only open access papers
`min_citations`	`50`	At least 50 citations
`max_citations`	`1000`	At most 1000 citations
`author_id`	`"A5023888391"`	Papers by a specific author
`institution_id`	`"I138006243"`	Papers from a specific institution
`source_id`	`"S137773608"`	Papers in a specific journal/source
`topic_id`	`"T10102"`	Papers on a specific topic

Example prompts:

"Find recent papers about transformer architectures in biology with over 100 citations"
"Search for open access reviews on CRISPR from 2022-2024"

`get_work` — Paper details

Fetch full metadata for a specific paper. Pass a DOI, PMID, or OpenAlex ID.

Parameters:

Param	Type	Required	Description
`identifier`	string	yes	DOI (`10.1234/example`), PMID (`29456894`), or OpenAlex ID (`W2741809807`)
`full`	boolean	no	Set `true` for complete raw metadata (default `false` returns trimmed summary)

Example prompts:

"Get details for DOI 10.7717/peerj.4375"
"What's the paper with PMID 29456894 about?"

`get_work_citations` — Who cites a paper

Find works that cite a given paper. Sorted by citation count by default — shows the most impactful citing works first.

Parameters:

Param	Type	Required	Description
`work_id`	string	yes	Paper identifier (OpenAlex ID or DOI)
`per_page`	number	no	Results per page (default 10)
`sort`	string	no	`cited_by_count:desc` (default) or `publication_date:desc`

`get_work_references` — What a paper cites

Explore the reference list of a paper — the foundation it builds on.

Parameters:

Param	Type	Required	Description
`work_id`	string	yes	Paper identifier
`per_page`	number	no	Results per page (default 10)

`get_works_by_author` — Author publications

List a researcher's publications with the same filter options as search_works.

Parameters:

Param	Type	Required	Description
`author_id`	string	yes	OpenAlex ID (`A5023888391`) or ORCID (`0000-0001-6187-6610`)
`filters`	object	no	Same filters as `search_works`
`sort`	string	no	`publication_date:desc` (default) or `cited_by_count:desc`
`per_page`	number	no	Results per page (default 10)

`search_authors` — Find researchers

Parameters:

Param	Type	Required	Description
`query`	string	yes	Author name
`institution_id`	string	no	Filter by institution
`has_orcid`	boolean	no	Only verified ORCIDs
`min_works`	number	no	Minimum publications
`min_citations`	number	no	Minimum total citations

`aggregate_works` — Statistics

Get distribution counts across any works query. Useful for understanding trends.

Parameters:

Param	Type	Required	Description
`query`	string	no	Optional search to scope results
`filters`	object	no	Same filters as `search_works`
`group_by`	string	yes	Field to group by (see options below)

Available group_by values:

publication_year — papers per year
type — by work type (article, review, preprint, etc.)
open_access.oa_status — by OA status (gold, green, hybrid, bronze, closed)
institutions.id — by institution
topics.id — by topic
primary_topic.field.display_name — by research field
primary_topic.domain.display_name — by domain
authorships.countries — by country
language — by language

Example prompt:

"How many AI papers were published each year since 2015?"

`set_config` — Runtime configuration

Update the API key or email for the current session without restarting.

Parameters:

Param	Type	Required	Description
`key`	string	yes	`"api_key"` or `"mailto"`
`value`	string	yes	The value to set (pass empty string to clear)

Response Format

All search tools return trimmed responses with only the fields useful for research:

{
  "meta": {
    "count": 12345,
    "page": 1,
    "per_page": 10
  },
  "results": [
    {
      "id": "https://openalex.org/W2741809807",
      "doi": "https://doi.org/10.7717/peerj.4375",
      "title": "The state of OA: a large-scale analysis...",
      "publication_year": 2018,
      "type": "journal-article",
      "authors": ["Heather Piwowar", "Jason Priem", "..."],
      "institutions": ["University of Washington", "..."],
      "source": "PeerJ",
      "cited_by_count": 1221,
      "oa_status": "gold",
      "topic": "scientometrics and bibliometrics research",
      "field": "Decision Sciences",
      "domain": "Social Sciences"
    }
  ]
}

Use get_work with full: true to get the complete raw response when you need all metadata.

Credit Costs

OpenAlex uses a credit system. Costs per request:

Operation	Credits
Single entity lookup (`get_work`, `get_author`)	0
List/filter query	1
Full-text search	10

Without an API key you get 100,000 credits/day. With a free key you get more headroom.

Supported ID Formats

The tools auto-detect identifier formats — no need to specify the type:

Format	Example
DOI	`10.7717/peerj.4375`
DOI URL	`https://doi.org/10.7717/peerj.4375`
PMID	`29456894`
OpenAlex ID	`W2741809807` (works), `A5023888391` (authors), `I138006243` (institutions)
ORCID	`0000-0001-6187-6610`
ROR URL	`https://ror.org/057zh3y96`

Architecture

mcp-openalex/
├── api-bridge-mcp/
│   ├── package.json       # Dependencies
│   ├── server.js          # MCP server entry — tool definitions and registration
│   ├── openalex.js        # OpenAlex HTTP client, ID resolver, filter builder, response trimmers
│   └── README.md          # Documentation
└── .gitignore

openalex.js handles all OpenAlex API communication, ID auto-detection, filter string construction, and response trimming
server.js registers 11 tools with the MCP SDK and wires them to the client
No database, no state — pure stateless API bridge over stdio

License

MIT

Name		Name	Last commit message	Last commit date
Latest commit History 4 Commits
.gitignore		.gitignore
README.md		README.md
openalex.js		openalex.js
package-lock.json		package-lock.json
package.json		package.json
server.js		server.js

Uh oh!

Folders and files

Latest commit

History

Repository files navigation

OpenAlex MCP Server

What is this?

Quick Start

Prerequisites

1. Clone or copy

2. Configure (optional but recommended)

3. Connect to your MCP client

Available Tools

Search & Discovery

Lookup & Detail

Citation Analysis

Author Publications

Analytics

Configuration

Tool Details & Examples

search_works — Find papers

get_work — Paper details

get_work_citations — Who cites a paper

get_work_references — What a paper cites

get_works_by_author — Author publications

search_authors — Find researchers

aggregate_works — Statistics

set_config — Runtime configuration

Response Format

Credit Costs

Supported ID Formats

Architecture

License

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

`search_works` — Find papers

`get_work` — Paper details

`get_work_citations` — Who cites a paper

`get_work_references` — What a paper cites

`get_works_by_author` — Author publications

`search_authors` — Find researchers

`aggregate_works` — Statistics

`set_config` — Runtime configuration

Packages