Connect OpenAlex to Meta Muse via MCP

💡 No install? Use cloud.anythingmcp.com directly. Sign in, click Connectors → OpenAlex, install it with no key, mint an MCP API key and you're done. No Docker, no git clone, no local server.

OpenAlex + Meta Muse

The open research catalog OpenAlex, no key needed: search and filter about 324 million scholarly works with their authors, institutions, journals, topics, publishers and funders, follow citations, check open access status and count records grouped by any field. Read-only; a free key raises the daily budget tenfold.

Prerequisites

See the full setup instructions baked into the connector (visible in the in-app store when you select the connector). The required environment variables for this connector are:

(none required; optional, recommended: OPENALEX_API_KEY)

Step 1 — Set up the connector

Works without a key: setup takes about a minute. The connector reads OpenAlex (https://api.openalex.org), the open catalog of the global research system run by the nonprofit OurResearch. Its data is CC0.

1. In the connector store, pick OpenAlex and install it. There is no Authorize with Provider step.
2. Recommended: keyless calls share one budget of 0.10 USD a day per network address, so on a shared server it runs out quickly. Create a free account at https://openalex.org/signup, copy the key from Settings > API key (https://openalex.org/settings/api-key) and paste it into OPENALEX_API_KEY. It is sent as Authorization: Bearer and gives 1 USD of free usage every day (ten times the keyless budget), 30 requests per second and cursor paging.
3. Check it with openalex_search_topics: it lists research topics with their hierarchy.

The old polite pool with mailto= was replaced by API keys in February 2026; the parameter is ignored.

What a call costs: single records (openalex_get_*) are free, a filtered list or a group count costs 0.0001 USD and a text search 0.001 USD. Every list answer reports its cost in meta.cost_usd.

Ids: OpenAlex ids are a letter plus digits: W (work), A (author), S (source), I (institution), T (topic), P (publisher), F (funder). The short form W2741809807 works everywhere, as does https://openalex.org/W2741809807. The get tools also take external ids: doi:10.7717/peerj.4375, orcid:0000-0003-1613-5981, ror:042nb2s44, issn:0028-0836, pmid:29456894. The AI never filters by a name: it resolves the name to an id first with openalex_autocomplete (quickest, free) or the matching search tool, then filters by that id.

What it reads:

• Works (articles, books, datasets, preprints, theses): search with search (keywords, AND, OR, NOT, "phrases"), search_exact (no stemming) or search_semantic (a paragraph, by meaning, at most 50 results), combined with filter and sort, for example cited_by_count:desc.
• Works of an author, institution, journal, topic or funder through filters such as authorships.author.id:A..., authorships.institutions.id:I..., primary_location.source.id:S..., topics.id:T... or funders.id:F..., plus publication_year:2020-2025, type:article or is_oa:true. Up to 100 DOIs at once with doi:10.1/a|10.2/b.
• One work in full: abstract, authors with institutions, source, topics, keywords, funders and awards, open access locations, references, related works and citations per year. Who cites it: a works search with cites:W....
• Authors (with ORCID, affiliations over time, h-index and i10-index), sources (journals, repositories, conferences with ISSN, open access and DOAJ status, APC price), institutions (with ROR, parents and children), the about 4,500 topics (domain > field > subfield > topic), publishers and funders, each as search and as single record.
• Counts with openalex_group_by: works per year, per open access status, per institution, per country or per topic, up to 200 groups, in one cheap call instead of paging.

Data formats: list tools return a slim record per result (id, names, counts, open access and author names); the matching get tool has everything else, and select keeps only the top-level fields you name. cited_by_count counts citing works; fwci is the field-weighted citation impact (1.0 is the world average). oa_status is diamond, gold, green, hybrid, bronze or closed. The abstract comes as abstract_inverted_index (each word with its positions), which the AI turns back into text. Dates are YYYY-MM-DD.

Paging: per_page up to 100 and page reach the first 10,000 results; meta.count is the total. Beyond that the AI uses cursor (* first, then meta.next_cursor), which needs an API key after the first page.

Read-only: every tool only reads, so there is nothing to confirm and no tool is switched off at install.

Limits: 429 means the daily budget is spent (without a key it is shared by everyone on the same address) or more than 10 requests per second were sent (30 with a key); semantic search allows 1 per second. Budgets reset at midnight UTC.

Common errors: 403 key_not_activated means the e-mail address of the key is not confirmed yet: open the activation link OpenAlex sent. 401 means a wrong key. 400 names an invalid filter or field, 404 an unknown id.

The connector does not download PDFs (0.01 USD each) and does not manage collections, saved searches, alerts or author profile corrections.

Step 2 — Install the adapter

curl -fsSL https://raw.githubusercontent.com/HelpCode-ai/anythingmcp/main/docker-compose.quickstart.yml -o docker-compose.yml
printf 'JWT_SECRET=%s\nENCRYPTION_KEY=%s\n' "$(openssl rand -hex 32)" "$(openssl rand -hex 32)" > .env
docker compose up -d

Step 3 — Add the connector in Muse

Your server URL: in AnythingMCP, open MCP Servers → the server this connector is on, and copy its URL (https://cloud.anythingmcp.com/mcp/…). Use it wherever this guide shows YOUR_SERVER_ID.

Muse adds remote MCP servers as custom connectors and signs in to them with OAuth.

1. In Muse, open Settings → Connectors and choose Add custom connector.
2. Paste the MCP server URL https://cloud.anythingmcp.com/mcp/YOUR_SERVER_ID.
3. Muse opens the AnythingMCP sign-in page: sign in and approve access.
4. In a chat, ask Muse for something that needs the OpenAlex tools.

You can also set it up from a chat, on the web, in the iOS and Android apps or in WhatsApp. Send Muse this message:

Create a Custom Connector for a new remote MCP server, then connect to it:
Name: AnythingMCP
Transport: remote streamable HTTP
URL: https://cloud.anythingmcp.com/mcp/YOUR_SERVER_ID
Auth: OAuth

Self-hosting? Use your instance's public HTTPS address instead of https://cloud.anythingmcp.com. Muse connects from Meta's cloud, so localhost does not work.

Available tools

Example prompts

• "How many works has ETH Zurich published per year since 2015, and what share is open access?"
• "Find the 20 most cited articles on perovskite solar cells since 2020, with DOI and journal."
• "Which institutions publish the most on CRISPR gene editing, and in which countries are they?"
• "Show the h-index, ORCID and main topics of this author, and list the works that cite their most cited paper."
• "Which funders appear most often on machine learning papers in medicine published in 2024?"

FAQ

Where is Muse available? Meta currently offers Muse in the US and Canada, on the web, in the iOS and Android apps and in WhatsApp.

Does Meta review this connector? No. Meta doesn't review custom connectors. Muse sees only the tools you switched on for this MCP server in AnythingMCP, and you can disconnect it anytime under Settings → Connectors.

Next steps

• Connector landing page
• Connect to Claude
• Connect to ChatGPT
• Connect to Cursor
• Connect to Gemini
• Connect to Microsoft Copilot