
Aiiq Mcp
io.github.shea256Published by GitHub account shea256v0.2.0Updated Oct 24, 2026
Query AI IQ (aiiq.org) model IQ, rankings, benchmarks, and methodology.
Overview
Read-only MCP server that lets an assistant query AI IQ (aiiq.org) model IQ scores, rankings, benchmarks, domains and methodology.
- What you can do
- List and inspect models with IQ, dimension scores, rank and cost, and compare several side by side
- Browse leaderboards: list available rankings and fetch ordered models for a ranking id
- Explore applied-capability domains and their benchmark leaderboards
- Look up the benchmark catalog and read how AI IQ is computed
- When to use it
- Useful when you want an assistant to look up or compare AI model IQ scores, rankings and benchmark results from AI IQ while discussing model choices, without leaving the chat.
- Requirements
- Node.js 18+ to run the npm package locally via npx; no API key required. Network access to the public AI IQ API over HTTPS. Optionally set AIIQ_API_BASE to point at a different versioned API root.
Installation
In SourceWeft
- Open Aiiq Mcp in the dashboard and add it to a workspace.
- Enable the server for the chats that should use its tools.
Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.
Other MCP clients
Follow the launch instructions in the repository.
README
@aiiq/mcp
Model Context Protocol server for AI IQ — query AI model IQ, rankings, applied-capability domains, benchmarks, and methodology from any MCP client (Claude Code, etc.).
Read-only. Talks to the public AI IQ API over HTTPS; no API key required.
Install
Requires Node.js 18+. No API key needed. Works in any MCP client.
Claude Code (one command):
Claude Desktop — edit claude_desktop_config.json (Settings → Developer → Edit Config) and add:
Then fully quit and reopen the app. (If Node is managed by nvm/asdf, use the absolute path to npx
as the command, since the desktop app doesn't inherit your shell PATH.)
Cursor / Windsurf / other clients — same JSON in the client's MCP config:
Tools
list_models— all public models with IQ, 7 dimension scores, emotional reasoning, rank, costget_model— full detail for one model (incl. per-benchmark results)list_rankings— available leaderboards with ids, names, model counts, and URLsget_ranking— ordered models for one ranking idlist_domains— applied-capability domains and benchmark countsget_domain— model composite IQs and benchmark leaderboards for one domainlist_benchmarks— benchmark catalogget_methodology— how AI IQ is computedcompare_models— side-by-side detail for several models
Config
AIIQ_API_BASE— override the versioned API base URL (defaulthttps://www.aiiq.org/api/v1). A custom value must be the API root that contains paths such as/modelsand/rankings.
Development
License
MIT
Source: README.md at commit d2b0b02
Tools
0Version history
1- v0.2.0LatestSep 16, 2026
Score
50% of the applicable checks met
How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.
- Reached in the last 30 days25 pts
Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.
- Auth declared10 pts
Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.
- Tools listed15 pts
Answer tools/list without a credential, or ship a package that starts in a sandbox without one.
- DescriptionRequired15 pts
- Logo10 pts
- Documentation15 pts
- License10 pts