openalex-mcp
The OpenAlex MCP server connects AI agents and CLI tools to OpenAlex, the world's largest free scholarly database (250M+ works, 300M+ authors, 100K+ institutions), enabling search, retrieval, and aggregation of academic data.
Core capabilities:
Search scholarly works: Find articles, books, datasets, preprints, and theses by keyword, filtered by year, type, institution, open-access status, citation count, and more.
Retrieve full work metadata: Get complete details for a single work via OpenAlex ID, DOI, PubMed ID, or MAG ID — including abstract, authors, affiliations, topics, and open-access URL.
Search and profile authors: Find researchers by name, institution, ORCID, or citation metrics; retrieve detailed profiles with h-index, i10-index, total citations, top topics, and year-by-year citation counts.
Search and explore institutions: Find universities and research organizations by name or country; get full details including ROR ID, country, type, h-index, and top research topics.
Discover publication sources: Search journals, conference proceedings, and repositories by name, OA status, or type; retrieve metadata including ISSN, publisher, DOAJ listing, and citation metrics.
Aggregate and analyze works: Group and count works by fields like publication year, type, language, or country to generate distribution statistics and trends.
Access filter reference: Consult a built-in filter syntax guide (
openalex://filter-reference) for constructing advanced queries.CLI support: Perform all operations directly from the terminal with JSON output support.
Broad AI client compatibility: Integrates with Claude Desktop, Claude Code, Cursor, VS Code Copilot, Windsurf, Zed, and other MCP-compatible clients.
Allows retrieval of scholarly work metadata by DOI using the OpenAlex API.
Enables GitHub Copilot to search and retrieve scholarly metadata from OpenAlex.
Allows searching for authors by ORCID identifier using the OpenAlex API.
Allows retrieval of scholarly work metadata by PubMed ID using the OpenAlex API.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@openalex-mcpfind papers about CRISPR gene editing from 2024"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
OpenAlex MCP Server & CLI
Connects any MCP-compatible AI agent — or your terminal — to OpenAlex: the world's largest free and open scholarly database with 250+ million works, 300+ million authors, 100,000+ institutions, and billions of citation links.
Works with: Claude Desktop · Claude Code · Cursor · VS Code Copilot · Windsurf · Zed · any MCP client — or standalone via the openalex CLI
No paywall. No institutional access required. CC0 licensed data.
Quickstart (2 minutes)
1. Get your free API key at openalex.org/settings/api — takes 30 seconds
2. Install the server
pip install openalex-mcp3. Add to your AI agent — copy your client's config from Configuration below.
Related MCP server: OpenAlex Author Disambiguation MCP Server
Prerequisites
Python 3.11+
python --version # needs 3.11 or higher
# Install if missing: https://python.org/downloadsuv (recommended)
# Windows
winget install astral-sh.uv
# macOS / Linux
curl -LsSf https://astral.sh/uv/install.sh | shNo uv? Use
pip install openalex-mcpinstead, then replaceuvx openalex-mcpin configs with the path fromwhere openalex-mcp(Windows) orwhich openalex-mcp(macOS/Linux).
Installation Options
Option A — uvx (recommended)
uvx openalex-mcp # install and run — uv handles everything
uv tool upgrade openalex-mcp # upgrade laterOption B — pip
pip install openalex-mcp
which openalex-mcp # macOS/Linux → /usr/local/bin/openalex-mcp
where openalex-mcp # Windows → C:\Users\YOU\AppData\...Option C — From source
git clone https://github.com/JOSETRA44/openalex-mcp.git
cd openalex-mcp
uv sync
cp .env.example .env # edit .env with your key
uv run openalex-mcp # run directlyEnvironment Variables
Variable | Required | Default | Description |
| Recommended | — | Free API key from openalex.org/settings/api |
| Alt to key | — | Your email — uses the "polite pool" (slower but free, no key) |
| No |
| Response cache duration in seconds (0 = disabled) |
| No |
| Retries on rate-limit errors |
| No |
|
|
Either
OPENALEX_API_KEYorOPENALEX_EMAILis required. The API key gives 10× more daily quota ($1/day free).
Configuration by Client
Replace YOUR_API_KEY_HERE with your actual key in all configs below.
Claude Desktop
Config file:
Windows:
%APPDATA%\Claude\claude_desktop_config.jsonmacOS:
~/Library/Application Support/Claude/claude_desktop_config.json
{
"mcpServers": {
"openalex": {
"command": "uvx",
"args": ["openalex-mcp"],
"env": {
"OPENALEX_API_KEY": "YOUR_API_KEY_HERE"
}
}
}
}Restart Claude Desktop after saving. You'll see a hammer icon (🔨) confirming tools are loaded.
Claude Code (CLI)
Add to your project's .mcp.json in the repo root:
{
"mcpServers": {
"openalex": {
"command": "uvx",
"args": ["openalex-mcp"],
"env": {
"OPENALEX_API_KEY": "YOUR_API_KEY_HERE"
}
}
}
}Or from source (local development):
{
"mcpServers": {
"openalex": {
"command": "uv",
"args": [
"--directory", "/absolute/path/to/openalex-mcp",
"run", "openalex-mcp"
],
"env": {
"OPENALEX_API_KEY": "YOUR_API_KEY_HERE"
}
}
}
}Verify with /mcp in the Claude Code prompt — you should see openalex with 9 tools.
Cursor
Config file: .cursor/mcp.json in project root, or globally:
Windows:
%APPDATA%\Cursor\User\globalStorage\cursor.mcp\mcp.jsonmacOS:
~/Library/Application Support/Cursor/User/globalStorage/cursor.mcp/mcp.json
{
"mcpServers": {
"openalex": {
"command": "uvx",
"args": ["openalex-mcp"],
"env": {
"OPENALEX_API_KEY": "YOUR_API_KEY_HERE"
}
}
}
}Settings → Features → MCP → toggle on → reload window.
VS Code + GitHub Copilot
Create .vscode/mcp.json in your workspace:
{
"servers": {
"openalex": {
"type": "stdio",
"command": "uvx",
"args": ["openalex-mcp"],
"env": {
"OPENALEX_API_KEY": "YOUR_API_KEY_HERE"
}
}
}
}Ctrl+Shift+P → GitHub Copilot: Configure MCP
Windsurf
Config file:
Windows:
%APPDATA%\Codeium\windsurf\mcp_config.jsonmacOS:
~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"openalex": {
"command": "uvx",
"args": ["openalex-mcp"],
"env": {
"OPENALEX_API_KEY": "YOUR_API_KEY_HERE"
}
}
}
}Zed
Edit ~/.config/zed/settings.json:
{
"context_servers": {
"openalex": {
"command": {
"path": "uvx",
"args": ["openalex-mcp"],
"env": {
"OPENALEX_API_KEY": "YOUR_API_KEY_HERE"
}
}
}
}
}Continue.dev
Edit .continue/config.yaml:
mcpServers:
- name: openalex
command: uvx
args:
- openalex-mcp
env:
OPENALEX_API_KEY: "YOUR_API_KEY_HERE"Any other MCP client (generic stdio)
command: uvx
args: ["openalex-mcp"]
env: OPENALEX_API_KEY=YOUR_API_KEY_HEREAutomated Setup Script
Auto-detect your installed clients and configure them:
# Interactive
python setup_mcp.py
# Non-interactive — configure all detected clients
python setup_mcp.py --key YOUR_API_KEY_HERE --yes
# Preview without writing files
python setup_mcp.py --key YOUR_API_KEY_HERE --dry-run --yes
# Show detected clients only
python setup_mcp.py --listThe script backs up existing config files before modifying them.
Available Tools (9 total)
Tool | What it does |
| Keyword + filter search over 250M+ scholarly works |
| Full metadata for one work (by OpenAlex ID, DOI, or PubMed ID) |
| Find researchers by name, institution, ORCID, or metrics |
| Author profile: h-index, i10-index, citations by year, topics |
| Search universities and research organizations by name/country |
| Institution details: country, type, h-index, top topics |
| Search journals, conferences, and repositories |
| Source details: ISSN, publisher, OA status, h-index |
| Group/count works by year, type, country, topic, etc. |
MCP Resource
Resource | Contents |
| Complete filter syntax: 100+ field codes, operators, and examples |
CLI (Command Line)
This package also installs an openalex command — a standalone terminal client for OpenAlex, no MCP client required. Every subcommand calls the exact same tool functions the MCP server uses (src/openalex_mcp/tools/*.py), so results match 1:1.
Install
pip install openalex-mcp # or: uv tool install openalex-mcp
openalex --help # confirm it's on your PATHFrom source: uv sync installs the openalex script into .venv/Scripts (Windows) or .venv/bin (macOS/Linux).
Commands
Command | Mirrors | What it does |
|
| Search works |
|
| Full work metadata |
|
| Search researchers |
|
| Full author profile |
|
| Search institutions |
|
| Full institution details |
|
| Search journals/venues |
|
| Full source details |
|
| Group/count works |
|
| Print the filter syntax reference |
Run openalex <command> --help for that command's flags. Every command supports --json for raw machine-readable output (pipe into jq, save to a file, script it), and --api-key / --email to override .env for a single call.
openalex search-works "federated learning privacy" -f "publication_year:>2021,type:article" -s "cited_by_count:desc"
openalex get-work 10.1038/s41586-021-03819-2
openalex search-authors "Geoffrey Hinton"
openalex aggregate-works publication_year -f "institutions.id:I865918315" --json | jq '.groups[0]'Auth and settings come from the same environment variables as the MCP server — see Environment Variables.
Example Queries
# Find recent papers on a topic
openalex_search_works(
query="federated learning privacy",
filters="publication_year:>2021,type:article",
sort="cited_by_count:desc"
)
# Get full paper details by DOI
openalex_get_work("10.1038/s41586-021-03819-2")
# Find an author and get their profile
openalex_search_authors(query="Geoffrey Hinton")
openalex_get_author("A2208157607")
# Find an institution, then see its recent output
openalex_search_institutions(query="UNAM", country_code="MX")
openalex_search_works(
filters="institutions.id:I865918315,publication_year:>2020",
sort="cited_by_count:desc"
)
# Annual publication trend for a topic
openalex_aggregate_works(
group_by="publication_year",
filters="type:article",
query="machine learning"
)
# Find open-access journals in a field
openalex_search_sources(
query="bioinformatics",
is_oa=True,
source_type="journal"
)Filter Quick Reference
Pattern | Example |
Year range |
|
Newer than |
|
Work type |
|
Open access only |
|
Highly cited |
|
By institution |
|
By country |
|
In a journal |
|
By language |
|
Papers citing a work |
|
Has DOI |
|
Has abstract |
|
For the full filter reference, read the openalex://filter-reference resource or see developers.openalex.org.
Rate Limits & Quotas
Auth method | Daily budget | Calls/second |
API key (free) | $1.00/day | ~10 |
Email (polite pool) | $0.10/day | ~3 |
No auth | $0.10/day | ~1 |
The server tracks rate-limit headers and automatically sleeps when the quota is low.
Action costs (free tier):
Action | Daily allowance |
Get single entity (by ID/DOI) | Unlimited |
List + filter | 10,000 calls / 1M results |
Full-text search | 1,000 calls / 100K results |
Tip: Use
openalex_get_work(unlimited) for single-record lookups rather than search.
Verify It's Working
MCP Inspector (interactive UI)
npx @modelcontextprotocol/inspector uvx openalex-mcpOpen the URL, click Connect, List Tools — you should see all 9 tools.
Quick smoke test
echo "" | OPENALEX_API_KEY=your_key uvx openalex-mcp
# Should start without errors, then exit when stdin closesRun unit tests (source install)
uv sync --group dev
uv run pytest tests/ -v
# Expected: 28 passedTroubleshooting
"Configuration error: Set OPENALEX_API_KEY..."
Add OPENALEX_API_KEY to your MCP client's env block. Get a free key at openalex.org/settings/api.
"command not found: uvx"
Install uv: docs.astral.sh/uv
401 error
API key is invalid. Regenerate at openalex.org/settings/api.
404 error on a work/author/institution
The identifier may be wrong. Try looking up the entity with a search tool first, then use the returned openalex_id.
"uvx openalex-mcp" slow on first run
uv is downloading and caching the package — subsequent starts are instant (~0.2s).
Tools appear but return empty results
Broaden your query: remove publication_year filters, or try query= instead of filters=.
vs. Scopus / Web of Science
OpenAlex | Scopus | Web of Science | |
Cost | Free | Subscription | Subscription |
Coverage | 250M+ works | 90M works | 170M works |
Open data | CC0 | Proprietary | Proprietary |
API quota | 10K calls/day | 20K/week | Limited |
Non-English | Excellent | Good | Fair |
Full-text links | Yes | Partial | Partial |
Project Structure
openalex-mcp/
├── src/openalex_mcp/
│ ├── server.py # FastMCP entry point (registers tools as MCP tools)
│ ├── cli.py # `openalex` CLI entry point (calls the same tool functions)
│ ├── cli_output.py # Human-readable table/detail rendering for the CLI
│ ├── config.py # Env var configuration (pydantic-settings)
│ ├── client.py # Async HTTP client + TTL cache + rate limiting
│ ├── exceptions.py # Error hierarchy
│ ├── formatters.py # Raw OpenAlex JSON → clean AI-friendly dicts
│ ├── tools/ # Core tool functions, shared by server.py and cli.py
│ └── resources/ # openalex://filter-reference static resource
├── openalex-researcher/ # Agent skill
│ ├── SKILL.md
│ └── references/ # Filter syntax, workflows, output reference
├── tests/ # Unit tests (28 cases, no network required)
├── setup_mcp.py # Automated config installer
├── .env.example # Environment variable template
└── pyproject.toml # Package definitionLicense
MIT — see LICENSE
Data from OpenAlex is CC0 (public domain).
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Servers
- AlicenseAqualityDmaintenanceEnables academic research through the OpenAlex API, allowing users to search for papers, authors, and institutions, retrieve citations, and fetch full-text content when available. Perfect for building intelligent research assistants that can explore academic literature and related works.Last updated87MIT
- AlicenseAqualityFmaintenanceEnables streamlined academic research and author disambiguation by providing AI agents with optimized access to the OpenAlex.org API. It supports searching for authors, resolving institutional affiliations, and retrieving scholarly works with detailed citation metrics.Last updated851MIT
- Alicense-qualityDmaintenanceEnables searching and retrieving scholarly works, authors, institutions, and citation networks from the OpenAlex catalog via natural language.Last updated22ISC
- Alicense-qualityDmaintenanceEnables AI agents to search and analyze OpenAlex scholarly database for OSINT research, including works, authors, institutions, funding, citations, and collaboration networks.Last updatedMIT
Related MCP Connectors
Access the OpenAlex academic research catalog — 270M+ publications.
Scholarly search: OpenAlex, Crossref, arXiv, OpenCitations and PubMed in one endpoint.
Search arXiv/Semantic Scholar/OpenAlex + medical evidence (PubMed/Europe PMC) + LaTeX/PDF tools.
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/JOSETRA44/openalex-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server