Skip to main content

MCP server for CEOInterviews — search 20,000+ executives, 1M+ quotes, and full verified interview transcripts via the Model Context Protocol

Project description

CEOInterviews MCP

Give your AI agent the world's largest verified database of executive interviews.

20,000+ executives · 1,000,000+ quotes · full transcripts · 4 MCP tools · 30-second install

PyPI Python MCP License: MIT Get API Key


Why this exists

Ask any AI agent "what has Jensen Huang said about AI infrastructure?" and you'll get hallucinated quotes, misattributed soundbites, and links to SEO-spam blogs. The web is full of fake CEO quotes — and LLMs eat them all.

CEOInterviews fixes this. Every transcript is AI- and human-verified to the original primary source: the actual interview, podcast, earnings call, or press conference. No fabrications. No misattributions. No "as an AI language model, here's a plausible-sounding quote."

❌ Other "executive quote" datasets ✅ CEOInterviews
Low-quality unverified source CEOInterviews AI-verified source
Unverified blog repost · No source URL · Misattributed quotes AI + human verified · Original primary source · Speaker-authenticated

This MCP server exposes that database to Claude, ChatGPT, Cursor, Windsurf, and any MCP-compatible agent — through 4 simple tools. Connect it once and your agent gains a research analyst with a perfect memory of what every CEO, CFO, central banker, and AI founder has actually said on the record.


30-second install

claude mcp add --transport http --scope user ceointerviews https://ceointerviews.ai/mcp \
  --header "Authorization: Bearer YOUR_API_KEY"

That's it. Open Claude Code, type /mcp, and you're connected. Grab a free API key →

Don't use Claude Code? Scroll to Connect any MCP client — Cursor, Windsurf, Claude Desktop, ChatGPT Agents, and raw HTTP all work.


The "aha" moment

Open your agent and ask any of these out loud:

"What has Jensen Huang said about AI infrastructure spending in the last 6 months?"

The agent calls get_transcripts(entity_name="Jensen Huang", keyword="AI infrastructure", filter_after_dt="...") and pulls back the actual NVIDIA earnings call transcripts and TIME interview, with source URLs and timestamps.

"Summarize Sam Altman's recent comments on Stargate and the $500B AI infra build-out."

Pulls Bloomberg interviews, OpenAI press conferences, and podcast appearances. Every claim is sourced — no hallucinations.

"What did Jerome Powell say about inflation and tariffs at the ECB forum?"

Hits search_quotes(entity_name="Jerome Powell", keyword="inflation tariffs", is_financial_policy=true). Returns the verified soundbites with the FOMC press conference URL.

"Find controversial statements from Elon Musk about tariffs."

search_quotes(entity_name="Elon Musk", keyword="tariffs", is_controversial=true). Speaker-verified, primary source.

This is what every "research GPT" promises and almost none deliver. The data layer has to actually exist — and be clean — first.


What you get

📈 20,000+ executives Every CEO, CFO, CTO across S&P 500, S&P 1500, NASDAQ 100, and global markets
💬 1,000,000+ quotes Tagged for notable, controversial, financial-policy sentiment
📜 Full interview transcripts Earnings calls, podcasts, conferences, press appearances — with source URLs and timestamps
🇺🇸 🇨🇳 🇪🇺 Global coverage US, China, and Europe-based companies. Public officials and Federal Reserve members. AI startup founders (OpenAI, Anthropic, Mistral, and hundreds more)
🤖 + 👤 AI + human verified Every transcript is checked for speaker authenticity. Source URLs and publication dates always included
Streamable HTTP MCP Hosted endpoint, Bearer token auth, no server to run

Pricing

$5 per 1,000 results. $0.005 per result. Billed weekly via Stripe. No monthly minimum, no commitment, no rate-limit gotchas.

Get your API key →


Connect any MCP client

Server URL https://ceointerviews.ai/mcp
Protocol Streamable HTTP (MCP)
Auth Authorization: Bearer YOUR_API_KEY

Claude Code (CLI + VSCode extension)

claude mcp add --transport http --scope user ceointerviews https://ceointerviews.ai/mcp \
  --header "Authorization: Bearer YOUR_API_KEY"

In VSCode, open the integrated terminal (Ctrl+backtick / Cmd+backtick) and run the same command. Then type /mcp in the Claude Code chat to verify.

Claude Desktop / Cursor / Windsurf

Paste this into your client's MCP config:

{
  "mcpServers": {
    "ceointerviews": {
      "type": "http",
      "url": "https://ceointerviews.ai/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY"
      }
    }
  }
}
Client Where to paste
Claude Desktop Settings → Developer → Edit Config → claude_desktop_config.json
Cursor Settings → MCP → Add Server
Windsurf Settings → MCP → Add Server
Claude Code (manual) Add to top-level mcpServers in ~/.claude.json

Anything else

Point your MCP client at the Streamable HTTP endpoint:

URL:    https://ceointerviews.ai/mcp
Auth:   Authorization: Bearer YOUR_API_KEY

The 4 tools

🎯 get_transcripts — your primary research tool

Returns full verified interview transcripts. Use this first for any question about what an executive has said.

Parameter Type Description
entity_name string Executive name, fuzzy matched (e.g. "Tim Cook")
entity_id int Exact entity ID from search_executives
company_name string Company name or ticker, fuzzy matched
company_id int Exact company ID
keyword string Search within transcript text (requires entity or company filter)
filter_after_dt string Only items after this date (ISO 8601)
filter_before_dt string Only items before this date (ISO 8601)
page_num int Page number (default 1)
page_size int Results per page, 1–500 (default 5). Keep small (5–7) — transcripts are large

🧑‍💼 search_executives

Find people. Returns name, title, institution, bio, and company metadata.

Parameter Type Description
keyword string Fuzzy search across name, title, company name, and ticker. Multi-term in any order: "Tim Cook", "AAPL", "CEO Tesla", "devinder kumar amd"
company_name string Filter by company name or ticker
gender string M, F, or O
is_snp500 bool Only S&P 500 company executives
is_nasdaq bool Only NASDAQ-listed company executives
is_snp1500 bool Only S&P 1500 company executives
is_nasdaq100 bool Only NASDAQ 100 company executives
is_ai_startup bool Only AI startup executives
is_top_startup bool Only top startup executives
is_usa_based bool Only US-based company executives
is_china_based bool Only China-based company executives
is_europe_based bool Only Europe-based company executives
is_public_company bool Only publicly traded company executives
page_num int Page number (default 1)
page_size int Results per page, 1–500 (default 10)

💬 search_quotes

Supplementary tool for short notable quotes. Use only when the user explicitly asks for soundbites or exact wording — for general research, prefer get_transcripts.

Parameter Type Description
entity_name string Executive name, fuzzy matched
entity_id int Exact entity ID
company_name string Company name or ticker, fuzzy matched
company_id int Exact company ID
keyword string Search within quote text
is_notable bool Only notable/important quotes
is_controversial bool Only controversial quotes
is_financial_policy bool Only financial/policy-related quotes
filter_after_dt string Only quotes after this date (ISO 8601)
filter_before_dt string Only quotes before this date (ISO 8601)
page_num int Page number (default 1)
page_size int Results per page, 1–500 (default 10)

🏢 search_companies

Look up companies and their classification flags.

Parameter Type Description
keyword string Company name or stock ticker
is_snp500 bool Only S&P 500 companies
is_nasdaq bool Only NASDAQ-listed companies
is_snp1500 bool Only S&P 1500 companies
is_nasdaq100 bool Only NASDAQ 100 companies
is_ai_startup bool Only AI startups
is_top_startup bool Only top startups
is_usa_based bool Only US-based companies
is_china_based bool Only China-based companies
is_europe_based bool Only Europe-based companies
is_public_company bool Only publicly traded companies
page_num int Page number (default 1)
page_size int Results per page, 1–500 (default 10)

Self-hosting (optional)

Most users should connect directly to the hosted server at https://ceointerviews.ai/mcp. Self-hosting only buys you the ability to run the MCP transport on your own infra — you'll still need a valid API key to call the upstream CEOInterviews REST API.

From PyPI

pip install ceointerviews-mcp
export CEOINTERVIEWS_API_KEY="your_api_key_here"
ceointerviews-mcp

From source

git clone https://github.com/ceointerviews/ceointerviews-mcp.git
cd ceointerviews-mcp
python -m venv venv && source venv/bin/activate
pip install -e .
cp .env.example .env  # then fill in your API key
ceointerviews-mcp

Docker

docker build -t ceointerviews-mcp .
docker run -e CEOINTERVIEWS_API_KEY=your_key -p 8001:8001 ceointerviews-mcp

Environment variables

Variable Default Description
CEOINTERVIEWS_API_KEY (none) Your CEOInterviews API key
CEOINTERVIEWS_API_BASE https://ceointerviews.ai API base URL
CEOINTERVIEWS_MCP_HOST 0.0.0.0 Server bind address
CEOINTERVIEWS_MCP_PORT 8001 Server port

Links


Built for the agents that have to be right.

Get an API key · Star on GitHub · MIT licensed

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

ceointerviews_mcp-0.1.0.tar.gz (10.4 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

ceointerviews_mcp-0.1.0-py3-none-any.whl (11.3 kB view details)

Uploaded Python 3

File details

Details for the file ceointerviews_mcp-0.1.0.tar.gz.

File metadata

  • Download URL: ceointerviews_mcp-0.1.0.tar.gz
  • Upload date:
  • Size: 10.4 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.11.4

File hashes

Hashes for ceointerviews_mcp-0.1.0.tar.gz
Algorithm Hash digest
SHA256 937b5e40b6c1d59d77f908a028fbc2c119852230b9cffbf738f644225a254fd9
MD5 87831d936b7cd59096b175aef8e878fa
BLAKE2b-256 e3b68a7c622bc3d2795b7d1a649ed6c9f03de572207e5d141cb76b67348de7a4

See more details on using hashes here.

File details

Details for the file ceointerviews_mcp-0.1.0-py3-none-any.whl.

File metadata

File hashes

Hashes for ceointerviews_mcp-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 5da5d9ee75c7f3955217e51c5b1db28be95ca695d17e1b4b7d8bd08081042f3f
MD5 b4d1b4f349bfe1bb650627125cfc3dba
BLAKE2b-256 09fb34085f6774c87d7e5e846ae27b7f955f08ba9dbf2f6ef24de0fd53313163

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page