Loading…
Install Meridian Consensus as an MCP server in Claude, Cursor, Claude Desktop, or any other AI tool that speaks the Model Context Protocol. Ask for a market size or competitive landscape in natural language and get committee-reviewed numbers cited back, with a link to the public preview.
MCP is an open standard that lets AI assistants — Claude, Cursor, ChatGPT, Cody, Continue.dev — connect to external tools and data sources through a uniform interface. Introduced by Anthropic in late 2024, it has become the de-facto way to give an AI model real-time access to a specific dataset without bolting on a custom API per client.
Think of it as USB-C for AI: one connector, many devices. Once a user installs the Meridian MCP server, every market-research query they make in their assistant is automatically routed through our committee-reviewed catalog — and the answer comes back with a citation link, not a hallucinated number.
A natural-language question becomes a cited answer in four steps. The AI client handles steps 01, 02 and 04; the Meridian MCP server handles step 03.
User asks
A buyer in Claude, Cursor, or ChatGPT types something like “how big is the cardiac stents market in EU by 2028”.
AI selects a tool
The assistant reads our tool descriptions and picks get_market_size with the right arguments.
Meridian responds
The MCP server queries the live preview catalog and returns a structured JSON payload — the figure, geography share, and a source_url.
One server URL — https://www.meridianconsensus.com/api/mcp — works across all MCP clients. Pick yours below.
claude mcp add meridian https://www.meridianconsensus.com/api/mcpConfirms the server is reachable, lists the 7 tools.
Settings → MCP → + Add new server. Paste:
https://www.meridianconsensus.com/api/mcpAvailable to the agent on every chat after enable.
Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):
{
"mcpServers": {
"meridian": {
"url": "https://www.meridianconsensus.com/api/mcp"
}
}
}Restart Claude Desktop. The Meridian tools appear under the 🔌 plug icon.
Any client that speaks the streamable-HTTP MCP transport works. URL:
https://www.meridianconsensus.com/api/mcpSeven read-only tools cover the full preview surface. The AI assistant picks the right one automatically; you do not call them by name.
search_marketsFree-text search across the published Meridian catalog. Returns up to 10 best matches with snippet and source_url.
Inputs
query· string (1–200 chars) · requiredIndustry name, technology, segment, or company.
When the AI calls this
Always called first when the user mentions a market by name ("cardiac stents", "GLP-1 agonists", "industrial robotics").
Example
// Input
{ "query": "cardiac stents" }
// Output (truncated)
{
"query": "cardiac stents",
"matches": [
{
"slug": "cardiovascular-stents",
"title": "Cardiovascular Stents",
"industry": "Medical Devices",
"snippet": "The cardiovascular stents market spans drug-eluting, bare-metal, and bioresorbable platforms…",
"tam_usd_millions": 9420,
"cagr_pct": 5.8,
"source_url": "https://meridianconsensus.com/markets/cardiovascular-stents"
}
]
}Real questions, the tool chain Meridian runs, and what the assistant says back.
You ask
"How big is the cardiovascular stents market in EU by 2028?"
Tool chain
get_market_size(slug='cardiovascular-stents', year=2028, geo='EU')Assistant answers
The EU cardiovascular stents market is projected at ~$3.24 B by 2028 (27.2 % of the global $10.94 B). Source: Meridian Consensus — meridianconsensus.com/markets/cardiovascular-stents.
You ask
"Who are the leading players in industrial robotics?"
Tool chain
search_markets(query='industrial robotics') → get_companies(slug=…)Assistant answers
The top vendors per Meridian are FANUC (18.4 % share), ABB (15.1 %), Yaskawa (13.7 %), KUKA (9.6 %), Mitsubishi Electric (8.2 %). Full top-25 with cap-stack is in the commissioned report. Source: meridianconsensus.com/markets/industrial-robotics.
What we log
Tool name and timestamp. Not the conversation, user identity, or IP. Aggregate counts feed our roadmap — which markets to deepen next.
Citation expectation
Every tool returns a source_url. We instruct the AI to present it. Free use requires attribution to Meridian Consensus when figures are surfaced to a third party.
Rate limits
None in v1. We may add per-IP throttling as adoption scales, with generous free quotas. Authenticated tiers (Q3 2026) will lift any caps.
What is the Model Context Protocol?
MCP is an open standard, created by Anthropic in 2024, that lets AI assistants like Claude, Cursor, and ChatGPT connect to external data sources and tools through a uniform interface. Once you install an MCP server, the AI can call its tools to fetch live data — and cite the source — without you having to copy-paste from a browser.
Is it free to use?
Yes. The Meridian MCP server exposes the same data that's already publicly readable on every /markets/[slug] preview page. There are no API keys, no rate limits in v1, and no account required. We may add authenticated tiers later (deeper data, write actions like commissioning a report), but the read endpoints remain free.
What data does it expose?
Headline TAM, CAGR, year-by-year forecast, top-5 companies with share, segment splits, regional shares, drivers, restraints, and scenario overlays for every published market. The full top-25 company list, all 30 chapters of analysis, and the per-claim provenance ledger remain in the commissioned report.
Does the AI always cite Meridian when it uses the data?
We instruct every tool to surface a source_url and ask the model to present it to the user. Compliance depends on the AI client and model — most do cite reliably, but it is not enforceable from our side. If you see a tool result presented without a Meridian citation, that's a behavior of the AI client, not the MCP server.
What do you log?
Tool name, anonymized request signature, and timestamp. We do not log user identity, IP, or AI-client conversation content. Logs are used to understand which markets are being asked about — that informs which markets we deepen next.
Commission a fresh 30-chapter intelligence brief on your specific market — same evidence framework, committee-signed, instant delivery. Once published, it also becomes queryable via this MCP server.
Assistant cites
The user sees the number alongside “Source: Meridian Consensus” and a link to the public preview.
Verify the install
Once installed, run this in your assistant to confirm the connection:
get_marketFull preview for one market by slug. Returns TAM, CAGR, forecast, top companies, segments, regions, drivers, restraints, and scenario overlay.
Inputs
slug· string · requirede.g. "cardiovascular-stents".
When the AI calls this
Called after search_markets identifies a slug, when the user wants depth on one market.
Example
// Input
{ "slug": "cardiovascular-stents" }
// Output (truncated)
{
"slug": "cardiovascular-stents",
"title": "Cardiovascular Stents",
"industry": "Medical Devices",
"base_year": 2026,
"forecast_year": 2030,
"market_size_usd_millions": 9420,
"forecast_size_usd_millions": 12480,
"cagr_pct": 5.8,
"top_companies": [ { "name": "Abbott", "share": 26.4 }, … ],
"segments": [ … ],
"regions": [ … ],
"drivers": [ … ],
"restraints": [ … ],
"source_url": "https://meridianconsensus.com/markets/cardiovascular-stents"
}get_market_sizeAnswers "how big is X" precisely. Optional year (uses yearly forecast) and geography (uses regional / country share).
Inputs
slug· string · requiredMarket slug.
year· integer · optionale.g. 2028. Defaults to base year.
geo· string · optionalRegion or country (e.g. "EU", "US", "APAC"). Defaults to global.
When the AI calls this
The most-frequent query type. Use when the user asks for a specific number.
Example
// Input
{ "slug": "cardiovascular-stents", "year": 2028, "geo": "EU" }
// Output
{
"slug": "cardiovascular-stents",
"title": "Cardiovascular Stents",
"size_usd_millions": 3240,
"currency": "USD",
"year": 2028,
"geography": "Europe",
"geography_share_pct": 27.2,
"cagr_pct": 5.8,
"source_url": "https://meridianconsensus.com/markets/cardiovascular-stents"
}get_companiesReturns the top companies in a market with HQ, revenue, and estimated share. Public preview returns the top-N; the full top-25 lives in the commissioned report.
Inputs
slug· string · requiredMarket slug.
When the AI calls this
When the user asks "who are the players" or "leading companies".
Example
// Input
{ "slug": "cardiovascular-stents" }
// Output
{
"slug": "cardiovascular-stents",
"title": "Cardiovascular Stents",
"companies": [
{ "name": "Abbott", "hq": "USA", "revenue": 43.7, "share": 26.4 },
{ "name": "Boston Scientific","hq": "USA", "revenue": 14.2, "share": 22.1 },
{ "name": "Medtronic", "hq": "IRL", "revenue": 31.5, "share": 18.6 }
],
"note": "Public preview shows top-5. Full top-25 with cap-stack and segment exposure is available in the commissioned report.",
"commission_url": "https://meridianconsensus.com/pricing?market=Cardiovascular%20Stents",
"source_url": "https://meridianconsensus.com/markets/cardiovascular-stents"
}get_forecastYear-by-year forecast for a market. Optional scenario (base / bull / bear) returns the scenario overlay.
Inputs
slug· string · requiredMarket slug.
scenario· "base" | "bull" | "bear" · optionalDefaults to base.
When the AI calls this
When the user asks for a multi-year projection or scenario sensitivity ("upside case", "downside").
Example
// Input
{ "slug": "cardiovascular-stents", "scenario": "bull" }
// Output (truncated)
{
"slug": "cardiovascular-stents",
"scenario": "bull",
"currency": "USD",
"forecast": [
{ "year": 2026, "value": 9420, "yoyGrowth": 5.2 },
{ "year": 2027, "value": 10100, "yoyGrowth": 7.2 },
{ "year": 2028, "value": 10940, "yoyGrowth": 8.3 }
],
"scenario_overlay": { "bull": { … }, "base": { … }, "bear": { … } },
"source_url": "https://meridianconsensus.com/markets/cardiovascular-stents"
}list_industriesReturns every industry with Meridian coverage, sorted by published market count. Zero-arg discovery tool.
Inputs
No arguments.
When the AI calls this
When the user is exploring what's available (no specific market mentioned).
Example
// Input
{}
// Output (truncated)
{
"industries": [
{ "industry": "Medical Devices", "market_count": 18 },
{ "industry": "Healthcare", "market_count": 14 },
{ "industry": "Industrial", "market_count": 9 }
],
"total": 24,
"catalog_url": "https://meridianconsensus.com/markets"
}list_markets_by_industryDrill into one vertical. Returns up to 50 markets in the given industry with size and link.
Inputs
industry· string · requirede.g. "Medical Devices".
When the AI calls this
After list_industries, when the user picks a vertical to explore.
Example
// Input
{ "industry": "Medical Devices" }
// Output (truncated)
{
"industry_query": "Medical Devices",
"markets": [
{ "slug": "cardiovascular-stents", "title": "Cardiovascular Stents", "tam_usd_millions": 9420, "cagr_pct": 5.8, … },
{ "slug": "insulin-pumps", "title": "Insulin Pumps", "tam_usd_millions": 6800, "cagr_pct": 9.1, … }
]
}You ask
"Give me a bull-case forecast for GLP-1 agonists."
Tool chain
get_forecast(slug='glp-1-agonists', scenario='bull')Assistant answers
Meridian's bull case projects the GLP-1 agonists market reaching ~$98 B by 2030 (vs $74 B base case), driven by accelerated obesity-indication adoption and earlier MASH approvals. Source: meridianconsensus.com/markets/glp-1-agonists.
You ask
"What does Meridian cover in medical devices?"
Tool chain
list_markets_by_industry(industry='Medical Devices')Assistant answers
Meridian covers 18 markets in Medical Devices, including cardiovascular stents ($9.4 B), insulin pumps ($6.8 B), surgical robotics ($5.1 B), and hearing aids ($9.3 B). Full catalog: meridianconsensus.com/markets.
Can I use this commercially in my product?
Yes, with attribution. The data is published under the same terms as the website: free to read and cite, paid to redistribute at scale or embed without attribution. See the editorial policy at /editorial-policy.
What happens if I ask about a market that isn't covered yet?
search_markets returns no matches and the AI will say so. You can commission a fresh 30-chapter brief on any market via the link in the answer — same evidence framework, instant delivery once you place the order.
What's the difference between this and your website?
The website is for humans browsing and reading. The MCP server is the same data restructured so AI assistants can call it as a tool inside another conversation — no tab-switching, no copy-paste, cited automatically.
Markets
Methodology
About
Research leadership