MCP.so
ログイン

Perplexity Sonar MCP Server

@felores

リポジトリ

Perplexity Sonar MCP Server について

MCP server for Perplexity API integration with Claude Desktop and other MCP clients

基本情報

カテゴリ

AI とエージェント

ライセンス

MIT

ランタイム

node

トランスポート

stdio

公開者

felores

設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "perplexity-sonar-mcp": {
      "command": "node",
      "args": [
        "--version",
        "#",
        "Should",
        "be",
        "18.0.0",
        "or",
        "higher"
      ]
    }
  }
}

ツール

ツールは検出されませんでした

ツールは README から自動的に抽出されます。メンテナーは ## Tools という見出しの下に記載することで、このタブに反映できます。

概要

What is Perplexity Sonar MCP Server?

An MCP server implementation that brings Perplexity’s Sonar models to Claude Desktop and other MCP clients. It enables real-time web search during conversations, giving Claude access to current information from the web.

How to use Perplexity Sonar MCP Server?

Configure Claude Desktop by editing claude_desktop_config.json with the npx command and your Perplexity API key as an environment variable. Alternatively, run in standalone mode with PERPLEXITY_API_KEY=your_key perplexity-sonar-mcp on port 3000 (or a custom PORT). The server exposes a single perplexity-chat tool with extensive parameters.

Key features of Perplexity Sonar MCP Server

  • Real-time web search during conversations
  • Multiple Sonar models for different needs
  • Rich parameter control (domains, recency, temperature, etc.)
  • Native integration with Claude Desktop
  • Standalone HTTP/SSE mode for other clients
  • Supports images and related questions in responses

Use cases of Perplexity Sonar MCP Server

  • Ask Claude for the latest news or developments on any topic
  • Perform in-depth research with comprehensive citations using sonar-deep-research
  • Limit search results to specific websites for authoritative sources
  • Filter results by recency (day, week, month) to get timely information
  • Generate creative or focused answers by adjusting temperature and parameters

FAQ from Perplexity Sonar MCP Server

Which models are available?

The server supports sonar (general), sonar-pro (enhanced), sonar-reasoning (logical analysis), and sonar-deep-research (in-depth with citations).

What are the requirements?

Node.js 18.0.0 or higher and a Perplexity API key from https://www.perplexity.ai/account/api.

How do I configure it for Claude Desktop?

Add the configuration to your Claude Desktop config file (~/Library/Application Support/Claude/claude_desktop_config.json on macOS, %APPDATA%\Claude\claude_desktop_config.json on Windows) with the npx -y @felores/perplexity-sonar-mcp command and PERPLEXITY_API_KEY environment variable.

Where can I find logs if something goes wrong?

On macOS, check ~/Library/Logs/Claude/mcp.log; on Windows, check %APPDATA%\Claude\logs\mcp.log.

Can I run the server standalone without Claude Desktop?

Yes, by setting the PERPLEXITY_API_KEY environment variable and running perplexity-sonar-mcp (or npx @felores/perplexity-sonar-mcp). It starts an HTTP/SSE server on port 3000 by default.

コメント

「AI とエージェント」の他のコンテンツ

🛡️ A.I.G（AI-Infra-Guard）

Tencent

A full-stack AI Red Teaming platform securing AI ecosystems via OpenClaw Security Scan, Agent Scan, Skills Scan, MCP scan, AI Infra scan and LLM jailbreak evaluation.

4.1K

LinkedIn MCP Server

stickerdaniel

Open-source MCP server for LinkedIn. Give Claude and any MCP-compatible AI agent access to profiles, companies, jobs, and messages.

2.7K

欢迎来到 智言平台

Shy2593666979

AgentChat 是一个基于 LLM 的智能体交流平台，内置默认 Agent 并支持用户自定义 Agent。通过多轮对话和任务协作，Agent 可以理解并协助完成复杂任务。项目集成 LangChain、Function Call、MCP 协议、RAG、Memory、HITL、Skill、Milvus 和 ElasticSearch 等技术，实现高效的知识检索与工具调用，使用 FastAPI 构建高性能后端服务。

797

Intervals.icu MCP Server

mvilanova

Model Context Protocol (MCP) server for connecting Claude and ChatGPT with the Intervals.icu API.

311

1Panel

1Panel-dev

🔥 1Panel is a modern, open-source VPS control panel — and the only one with native AI agent support. Run Ollama models, deploy OpenClaw agents, and manage your entire server stack from one clean web interface.

36.1K