Skip to main content
Glama
perplexity-mcp-server

perplexity-mcp-server

by thierrysays

perplexity-mcp-server

MCP server for structured competitive/market-intelligence veille workflows on top of the Perplexity Sonar API.

Correction (2026-08-26): an earlier version of this README claimed no native Perplexity MCP connector existed anywhere. That was wrong — Perplexity publishes an official MCP server (github.com/perplexityai/modelcontextprotocol), including a Perplexity-hosted remote endpoint at https://api.perplexity.ai/mcp that's reachable from a cloud-scheduled session with zero self-hosting. If you just need generic ad hoc Sonar search, register that instead of building or running anything here.

What the official server doesn't give you is a structured, comparable weekly veille output: it exposes generic search/reasoning (Agent API, Search API), not a fixed-section company brief or a JSON-schema-constrained signal sweep. That's this server's actual reason to exist — it's a thin, narrow layer of two workflow tools, not a replacement for Perplexity's own MCP server.

Deploy this once (see "Remote deployment" below) with PERPLEXITY_API_KEY set as a platform secret, and any MCP client — including a cloud-scheduled Cowork/Claude session — can call it over HTTP without ever touching the local machine or a device bridge. Register it alongside Perplexity's official server, not instead of it.

Tools

  • perplexity_company_news — structured company news brief (governance, executives, strategic events, financial situation, senior HR moves), fixed section headings, lookback_days instead of free-text date ranges. Output language is fr (default) or en via the language parameter — headings switch accordingly (e.g. Gouvernance/Governance, Dirigeants/Executives).

  • perplexity_market_signals — sector-wide signal sweep (M&A, funding, appointments, restructurings) constrained to a JSON schema server-side, so the result is a real array of discrete signals, not markdown to re-parse. Returns an empty array (not an error) when nothing dated and reliable is found for the period — that's a meaningful result, not a failure.

Both tools support response_format: "markdown" | "json" and return citations/sources. For general-purpose search, use Perplexity's official MCP server instead — see the correction above.

Related MCP server: INDUSS Research Intelligence MCP Server

Setup

npm install
cp .env.example .env   # then set PERPLEXITY_API_KEY
npm run build

Local use (stdio) — quick start, does NOT fix the cloud-scheduling problem

Good for testing, or for any workflow that already runs on your own machine. Register it in your MCP client's config (e.g. Claude Desktop's claude_desktop_config.json):

{
  "mcpServers": {
    "perplexity": {
      "command": "node",
      "args": ["/absolute/path/to/perplexity-mcp-server/dist/index.js"],
      "env": { "PERPLEXITY_API_KEY": "pplx-..." }
    }
  }
}

This still ties availability to that machine being on — it does not solve the original problem of a cloud-scheduled task needing the key. For that, use the HTTP deployment below.

Remote deployment (streamable HTTP) — the actual fix for cloud-scheduled tasks

Run the same code with TRANSPORT=http, on any Node-capable host (a small VPS, Render, Railway, Fly.io, a container on any cloud provider). The only requirements are: Node 18+, the ability to set an environment variable secret, and a public HTTPS URL.

PERPLEXITY_API_KEY=pplx-... TRANSPORT=http PORT=3000 npm start

The server exposes:

  • POST /mcp — the MCP endpoint (streamable HTTP, stateless JSON — a fresh server+transport per request, so nothing leaks between calls and it scales horizontally without shared state)

  • GET /health — plain liveness check for your hosting platform's health probe

Once deployed behind HTTPS (https://your-host/mcp), register that URL as a remote MCP connector in the workspace so the Perplexity-dependent scheduled task can reach it directly, with the key living server-side — the same operational model Pappers already uses (native connector, no local secret in the credential path).

What has been verified in this build

  • npm run build compiles cleanly (strict TypeScript, no any).

  • Both transports start without crashing (stdio and HTTP).

  • GET /health responds correctly.

  • Full MCP handshake (initialize, tools/list) returns both tools with correct JSON schemas, descriptions, and annotations.

  • tools/call with an invalid API key returns a clean, actionable isError: true result (401 guidance pointing at the Perplexity dashboard) — the server does not crash and stays responsive afterward.

What has NOT been verified yet

No real PERPLEXITY_API_KEY was available in the build environment, so an actual successful Sonar call — real answer content, real citations, the market_signals JSON-schema-constrained response actually parsing as valid JSON — has not been exercised end-to-end. Before relying on this in the production veille routine, run each of the two tools once with a real key (locally via stdio is enough for this check) and confirm the output matches the shapes documented above.

Cost note

perplexity_company_news and perplexity_market_signals both default to the sonar-pro model (deeper search, more sources, higher cost per call than sonar) because the veille use case prioritizes completeness and citation quality over latency/cost — this is the deliberate default for both of this server's tools, confirmed 2026-08-26. There is no perplexity_search tool here to opt into a cheaper model for one-off queries; use Perplexity's official MCP server for that (see the correction at the top of this file). Review Perplexity's current per-model pricing before running this at the weekly-routine scale across dozens of companies — a full run enriches on the order of 50-90 companies plus one market-signals sweep, so the marginal cost of sonar-pro vs sonar compounds quickly.

Available Tools

2 tools
perplexity_company_newsPerplexity Company News BriefA
Read-only

Get a structured, cited news brief on one company: governance, executives, strategic events, financial situation, and senior HR moves.

This is a workflow tool built on top of Perplexity's Sonar chat/completions API: it fixes the prompt structure and section headings so every call returns a comparable, five-section brief, and maps a plain lookback_days number to Perplexity's recency buckets (there is no native "3 months" filter, so lookback_days=90 is mapped to the nearest bucket, 'month').

Args:

  • organization (string): Company name, e.g. "Doctolib".

  • lookback_days (number, 1-365, default 90): how far back to search ("90" = "last 3 months").

  • language ('fr'|'en'): default 'fr'.

  • response_format ('markdown'|'json'): default 'markdown'.

Returns: Markdown: five headed sections (Gouvernance/Governance, Dirigeants/Executives, Événements stratégiques/Strategic events, Situation financière/Financial situation, Mouvements RH/Senior HR moves), each stating "Rien à signaler"/"Nothing to report" if empty, followed by numbered Sources. JSON: { "answer": string (same structured text), "citations": string[], "search_results": [...], "organization": string, "lookback_days": number }

Examples:

  • Use when: "Quelle est l'actualité récente de Nexans ?" -> organization="Nexans"

  • Use when: weekly veille loop over a list of target companies -> call once per company

  • Don't use when: you need sector-wide signals across many companies -> use perplexity_market_signals instead.

Error Handling:

  • Returns "Error: ... Invalid Perplexity API key" if PERPLEXITY_API_KEY is missing/invalid (401).

  • Returns "Error: ... rate limit exceeded" if too many requests (429).

Inputs: organization (required), language, lookback_days, response_format.

perplexity_market_signalsPerplexity Market SignalsA
Read-only

Find recent, dated market signals (M&A, funding rounds, executive appointments, restructurings, etc.) relevant to a given professional profile and region, as a structured list.

Unlike a free-text Sonar query, this tool constrains Perplexity's response to a JSON schema server-side, so the result is a real array of discrete signals — not markdown you have to re-parse. It returns an empty list rather than fabricated signals when nothing dated and reliable is found; that empty list is itself a meaningful, actionable result (it means the sweep ran and found nothing this period), not a failure.

Args:

  • profile (string): Target professional profile, e.g. "CTO groupe / VP Technology / Chief Transformation Officer / Group CIO".

  • regions (string[], default ["France","Europe","Luxembourg"]): geographic scope.

  • lookback_days (number, 1-30, default 7): how far back to search.

  • signal_types (string[], default M&A/levée de fonds/nomination/restructuration): categories.

  • language ('fr'|'en'): default 'fr'.

  • response_format ('markdown'|'json'): default 'markdown'.

Returns: Markdown: one numbered section per signal (title, type, date, source, summary), or an explicit "no signal found" statement if the array is empty. JSON: { "signals": [{title,type,date,source_url,summary}], "count": number, "lookback_days": number, "regions": string[] }

Examples:

  • Use when: weekly sector-wide sweep independent of any specific tracked company

  • Don't use when: you already know the company and want its own news -> use perplexity_company_news instead.

Error Handling:

  • Returns "Error: ... Invalid Perplexity API key" if PERPLEXITY_API_KEY is missing/invalid (401).

  • Returns "Error: ... rate limit exceeded" if too many requests (429).

  • If Perplexity returns non-JSON content despite the schema constraint, returns "Error: could not parse signals JSON" with the raw answer included so nothing is silently lost.

Inputs: profile (required), regions, language, signal_types, +2 more.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A4.8/5.0

    Scored across 2 tools

    Disambiguation5/5

    The two tools have cleanly separated domains: one compiles a structured news brief for a single named company, while the other scans for market-wide signals across profiles and regions. Their descriptions explicitly cross-reference when not to use each, removing ambiguity.

    Naming Consistency5/5

    Both tool names follow the same `perplexity_<object>` snake_case pattern and clearly indicate their target: `company_news` vs `market_signals`. No mixed conventions or vague verbs.

    Tool Count3/5

    At two tools, the surface is on the thin side and borders on feeling minimal for a server named after Perplexity. Each tool does earn its place and covers a distinct workflow, but the count is at the low end of acceptable.

    Completeness4/5

    The pair covers the two main read-oriented intelligence workflows: focused company monitoring and broader market sweeps. A general free-text search or ask tool is missing, but the server's stated purpose appears to be these richer structured workflows, so the gap is minor.

    Maintenance

    ActivityInactive
    ResponsivenessOutdated data

    Related MCP Connectors

    Related MCP Servers

    • F
      license
      A
      quality
      D
      maintenance
      Provides an institutional research backend for AI assistants, with 15 tools for company, financial, funding, competitor, industry, and news intelligence, plus Markdown/PDF report generation, featuring deterministic source routing, extraction, validation, and citation generation.
      15
      -
    • F
      license
      A
      quality
      C
      maintenance
      Enables AI assistants to perform institutional research and due diligence by retrieving, extracting, validating, and citing data from public sources, then generating reports in markdown, HTML, or PDF.
      23
      -
    • A
      license
      A
      quality
      B
      maintenance
      Enables agents to perform cited social listening and company research, returning mentions, sentiment, and themes in a single low-latency USDC-paid call.
      12
      1
      MIT