Perplexity Sonar
БесплатноНе проверенAn MCP server that brings Perplexity's Sonar models with real-time web search capabilities to Claude Desktop and other MCP clients.
Описание
An MCP server that brings Perplexity's Sonar models with real-time web search capabilities to Claude Desktop and other MCP clients.
README
An MCP server implementation that brings the power of Perplexity's Sonar models to Claude Desktop and other MCP clients. Unlike traditional MCP servers, this implementation provides access to Perplexity's advanced AI models with web search capabilities, allowing Claude to perform real-time web searches and provide up-to-date information during conversations.
Why Perplexity Sonar MCP?
- Real-time Web Search: Access current information from the web during conversations
- Multiple Models: Choose from various Sonar models to fit your needs:
sonar: Balanced model for general usesonar-pro: Enhanced capabilities for complex taskssonar-reasoning: Specialized in logical reasoning and analysissonar-deep-research: In-depth research with comprehensive citations
- Rich Parameters: Fine-tune your queries with extensive parameters:
- Control search domains and recency
- Adjust response format (markdown/JSON)
- Manage temperature and token generation
- Get related questions and images
- Seamless Integration: Works natively with Claude Desktop and other MCP clients
Requirements
- Node.js 18.0.0 or higher
- A Perplexity API key (get one at https://www.perplexity.ai/account/api)
- Claude Desktop (for desktop integration)
To verify your Node.js installation:
node --version # Should be 18.0.0 or higher
If Node.js is not installed or needs updating:
- Download from nodejs.org
- Follow the installation instructions for your operating system
- Restart your terminal after installation
Usage
With Claude Desktop
Edit your Claude Desktop configuration file:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
- macOS:
Add the following configuration:
{ "mcpServers": { "perplexity": { "command": "npx", "args": [ "-y", "@felores/perplexity-sonar-mcp" ], "env": { "PERPLEXITY_API_KEY": "your_api_key_here" } } } }Restart Claude Desktop
Standalone Mode
You can also run the server in standalone mode with HTTP/SSE support:
PERPLEXITY_API_KEY=your_api_key_here perplexity-sonar-mcp
This will start the server on port 3000 by default. You can specify a different port:
PERPLEXITY_API_KEY=your_api_key_here PORT=8080 perplexity-sonar-mcp
API Reference
The server exposes the following MCP tool:
perplexity-chat
Parameters:
model(optional): Model to use (default: "sonar")- Options:
sonar: Balanced model for general usesonar-pro: Enhanced capabilities for complex taskssonar-reasoning: Specialized in logical reasoningsonar-deep-research: In-depth research with citations
- Options:
messages: Array of message objects with:role: "system" | "user" | "assistant"content: string
temperature(optional): Controls randomness (0-2)- Lower values (e.g., 0.3) for more focused responses
- Higher values (e.g., 0.8) for more creative responses
max_tokens(optional): Maximum tokens to generatetop_p(optional): Controls diversity via nucleus sampling (0-1)search_domain_filter(optional): Array of domains to limit citations to- Example: ["wikipedia.org", "github.com"]
return_images(optional): Whether to return images from search resultsreturn_related_questions(optional): Whether to return related questionssearch_recency_filter(optional): Filter search results by recency- Options: "month", "week", "day", "hour"
top_k(optional): Maximum number of search results to considerstream(optional): Whether to stream the responsepresence_penalty(optional): Penalizes token presencefrequency_penalty(optional): Penalizes token frequencyoutput_format(optional): Response format ("json" or "markdown", default: "markdown")
Example Queries
Here are some examples of what you can ask Claude using this server:
Basic search with default model: "What are the latest developments in quantum computing?"
Deep research with citations: "Using sonar-deep-research, explain the current state of fusion energy."
Domain-specific search: "Search only on arxiv.org and nature.com for recent papers about large language models."
Recent news search: "What happened in AI research this week? Use search_recency_filter: 'week'"
Development
Clone the repository:
git clone https://github.com/felores/perplexity-sonar-mcp.git cd perplexity-sonar-mcpInstall dependencies:
npm installBuild the project:
npm run buildRun in development mode:
npm run dev
Troubleshooting
If you encounter issues:
- Verify your API key is correct
- Check Node.js version compatibility
- Look for errors in Claude Desktop logs:
- macOS:
~/Library/Logs/Claude/mcp.log - Windows:
%APPDATA%\Claude\logs\mcp.log
- macOS:
- Try running in standalone mode to isolate issues
License
MIT
Установка Perplexity Sonar
У этого сервера нет опубликованного пакета — он собирается из исходников. Открой репозиторий и следуй инструкции в README.
▸ github.com/felores/perplexity-sonar-mcpFAQ
Perplexity Sonar MCP бесплатный?
Да, Perplexity Sonar MCP бесплатный — установка в пару кликов через Unyly без оплаты.
Нужен ли API-ключ для Perplexity Sonar?
Нет, Perplexity Sonar работает без API-ключей и переменных окружения.
Perplexity Sonar — hosted или self-hosted?
Доступен hosted-вариант: Unyly запускает сервер в облаке, локальная установка не обязательна.
Как установить Perplexity Sonar в Claude Desktop, Claude Code или Cursor?
Открой Perplexity Sonar на unyly.org, выбери вкладку своего клиента (Claude Desktop, Claude Code, Cursor) и нажми Install — конфиг сгенерируется автоматически, без правки JSON.
Похожие MCP
Fetch
Web content fetching and conversion for efficient LLM usage.
AWS KB Retrieval
Retrieval from AWS Knowledge Base using Bedrock Agent Runtime.
автор: modelcontextprotocolSpring AI MCP Server
Provides auto-configuration for setting up an MCP server in Spring Boot applications.
llm-analysis-assistant
A very streamlined mcp client that supports calling and monitoring stdio/sse/streamableHttp, and can also view request responses through the /logs page. It also
автор: xuzexin-hzCompare Perplexity Sonar with
Не уверен что выбрать?
Найди свой стек за 60 секунд
Автор?
Embed-бейдж для README
Похожее
Все в категории ai
