Name: AI Authority Index: live data (API, feeds, and MCP server)
Creator: Jeremy Liddle
License: https://creativecommons.org/licenses/by/4.0/

AI Authority Index data: API, feeds, and MCP server

Third Hemisphere's AI Authority Index measures how the AI engines buyers ask describe, recommend, and source the companies in each market. Its editions, every market's Standing Reports, and each company's CRISP score are published here as live data, free to read and reuse under CC BY 4.0.

The static copies on Zenodo, Hugging Face, Kaggle, and GitHub are snapshots of one edition. These services are the current record: they change the moment an edition publishes or a market is measured again, so anything built on them stays right.

Start in one line

Every published edition

curl https://thirdhemisphere.agency/api/public/v1/editions

An edition's results

curl "https://thirdhemisphere.agency/api/public/v1/editions/personal-lending/results?section=personal-lending"

Find a company, then read its scores

curl "https://thirdhemisphere.agency/api/public/v1/companies?q=credit"

With a key, for the higher limits

curl -H "x-api-key: YOUR_KEY" https://thirdhemisphere.agency/api/public/v1/markets

Endpoints

Every endpoint is a GET under https://thirdhemisphere.agency/api/public/v1, answers JSON, and is described in the OpenAPI 3.1 spec.

/api/public/v1/editions: Every published edition, with its sections, dates, DOIs, and links.
/api/public/v1/editions/{slug}: One edition.
/api/public/v1/editions/{slug}/results?section=: One section's results: each company's Star Chart cell, CRISP score, factor readings, and answers naming it.
/api/public/v1/markets: Every market measured, with its count of Standing Reports and when it was measured.
/api/public/v1/markets/{market}/standing: A market's companies, most named first. Takes limit and offset.
/api/public/v1/standing/{market}/{company}: One Standing Report: the headline score, the five factors, presence, and the date.
/api/public/v1/companies?q=: Search companies by name.
/api/public/v1/companies/{company}: One company across every market it is measured in.
/api/public/v1/companies/{company}/history: Every reading published for a company, oldest first.

Every answer says where it came from

Each response wraps its data with the fields a reuse needs to credit it: attribution, citation, licence, doi (once an edition's Zenodo record has one), canonical_url (the public page the data belongs to), and updated_at. A context block says what the Index is, how CRISP measures it, who wrote it, and who publishes it.

Quote a score exactly as returned, with its market, its month, and "Third Hemisphere's AI Authority Index", and link canonical_url. Source: Third Hemisphere's AI Authority Index (https://thirdhemisphere.agency/ai-authority-index), CC BY 4.0.

Feeds

New editions, new markets, and each market's updated scores, as they publish. One stream in three formats:

MCP server for AI agents

A read-only remote MCP server at https://thirdhemisphere.agency/mcp, on the Streamable HTTP transport, with no sign-in. Every tool result ends with its citation and link, so an agent that answers from it credits the Index. Tools: list_index_editions, get_edition_results, get_company_authority_score, get_company_score_history, search_companies, list_markets, get_market_standing.

Claude

In Claude (web or desktop), open Settings, then Connectors, and choose Add custom connector.

Name it "Third Hemisphere AI Authority Index" and paste https://thirdhemisphere.agency/mcp as the URL. No sign-in is needed.

In Claude Code, run the command below.

claude mcp add --transport http ai-authority-index https://thirdhemisphere.agency/mcp

ChatGPT

In ChatGPT, open Settings, then Apps and Connectors, and turn on developer mode under Advanced.

Create a connector, paste https://thirdhemisphere.agency/mcp as the MCP server URL, and choose no authentication.

Cursor

Add the server to .cursor/mcp.json in your project, or to ~/.cursor/mcp.json for every project.

{
  "mcpServers": {
    "ai-authority-index": {
      "url": "https://thirdhemisphere.agency/mcp"
    }
  }
}

Any other MCP client

Use the Streamable HTTP transport at https://thirdhemisphere.agency/mcp. Send a key in the x-api-key header for the higher limits.

Access and limits

Reading is free. Limits keep the service fast for everyone, and every tier carries the same attribution.

Open: 30 requests a minute, 1,000 a day. No key. Every read, at a pace that suits a person, a site, or an agent answering a question.
Free key: 120 requests a minute, 20,000 a day. A free key sent to your email. Higher limits for sites and tools that sync the data.
Pro: 600 requests a minute, 500,000 a day. For platforms and research teams that sync every market: the highest limits, whole-market pages in one call, and support.

Get a free key

A free key raises your limits. It is emailed to the address you give, and only there, so check your inbox. Send it in the x-api-key header, or as a Bearer token, to the API and the MCP server. You can also ask for one with a POST to https://thirdhemisphere.agency/api/public/v1/keys.

Pro access and data partnerships

For platforms, agencies, research teams, and AI products that sync every market: the highest limits, whole markets in one call, the full score history, early access to new markets, and a named contact. The data stays CC BY 4.0, with the same attribution.

Talk to us about Pro access

For crawlers and agents

How to cite

Credit Third Hemisphere and link the page the data came from. For an edition: the citation its record carries, with its DOI. For a score: the company, the market, the month, and "Third Hemisphere's AI Authority Index". Licence: Creative Commons Attribution 4.0 International (CC BY 4.0), https://creativecommons.org/licenses/by/4.0/.

Using the AI Authority score

AI Authority Index data: API, feeds, and MCP server