Model Context Protocol

Use AdmitBase inside your AI assistant

AdmitBase ships a public MCP server. Any MCP-capable client — Claude Desktop, Cursor, Claude.ai connectors, ChatGPT custom connectors — can query admissions data for 500+ professional schools, calculate match scores, and compare applicant outcomes. Three of the five tools work anonymously; no signup required.

Public endpoint

Copy this URL into any MCP client that supports remote streamable-HTTP servers.

https://admitbase.com/api/mcp

Transport: streamable-http · Protocol: MCP 2025-06-18 · No authentication required for read-only tools · 60 requests/min/IP anonymous, 600/min authenticated.

What you can do

search_schoolsPublic

Find schools by program (law, medical, dental, MBA, pharmacy, veterinary, optometry), name, or ranking range.

get_school_statsPublic

Full admissions stats: GPA/test percentiles, acceptance rate, tuition, employment outcomes.

calculate_match_scorePublic

Given a user's GPA + test score, return Safety/Target/Reach/Far-Reach for any or every school in a program, plus admission probability.

save_my_statsAuth required

Persist an applicant's stats under their API key so later calls can reference them.

compare_to_applicantsAuth required

Show how a user's GPA + test score rank against anonymous community outcome reports (accepted / waitlisted / rejected).

Install in 30 seconds

Choose your AI assistant. Anonymous access works out of the box; authenticated tools need a key from your profile.

Claude Desktop

Edit claude_desktop_config.json (Settings → Developer → Edit Config) and add the mcpServers entry:

{
  "mcpServers": {
    "admitbase": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://admitbase.com/api/mcp"]
    }
  }
}

Restart Claude Desktop. The five AdmitBase tools appear in the connectors panel. To unlock authenticated tools, swap in:

{
  "mcpServers": {
    "admitbase": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://admitbase.com/api/mcp",
        "--header",
        "Authorization: Bearer amb_YOUR_KEY_HERE"
      ]
    }
  }
}

Cursor

Create or edit .cursor/mcp.json in your project (or ~/.cursor/mcp.json for global access):

{
  "mcpServers": {
    "admitbase": {
      "url": "https://admitbase.com/api/mcp"
    }
  }
}

ChatGPT Custom Connector

1. Open ChatGPT → Settings → Connectors → Add custom connector.

2. Paste https://admitbase.com/api/mcp as the server URL.

3. Leave auth blank for anonymous use, or paste an amb_key as a Bearer token for personalised tools.

Any other MCP client

POST JSON-RPC 2.0 requests to https://admitbase.com/api/mcp. Discovery manifest at /.well-known/mcp.json.

Methods supported: initialize, tools/list, tools/call, ping.

Optional: personalised tools

The two personalised tools — saving an applicant's stats and comparing them to community outcomes — require an API key tied to an AdmitBase account. Generate one in your profile, then paste it as a Bearer token in your MCP client.

Generate an API key

Try a prompt

Once installed, ask your assistant something like:

“Use AdmitBase to find the top 20 medical schools and tell me which of them my 3.7 GPA and 515 MCAT would be a target for.”
“What are NYU Law's 25/50/75th LSAT and GPA numbers? Pull from AdmitBase.”
“Compare a 3.85 GPA / 168 LSAT applicant against all T-14 law schools. Use AdmitBase's match calculator.”

Want the full story?

The long-form article covers what MCP is, why we built this, the data sources behind each tool, rate limits, and the roadmap.

Read: AdmitBase MCP for AI assistants