Parallect MCP Server

Parallect MCP Server

Enables AI assistants to perform multi-provider deep research via Parallect.ai, including submitting queries, tracking progress, and retrieving synthesized reports with citations and evidence chains.

Category
访问服务器

README

Parallect MCP Server

A Model Context Protocol server for Parallect.ai — multi-provider deep research from AI.

This MCP server gives AI assistants (Cursor, Claude Desktop, Claude Code, etc.) access to Parallect's deep research capabilities: submit queries, track progress, retrieve synthesized reports with cross-referenced citations, and explore claims with full evidence chains.

Quick Start

Cursor

Add to your Cursor MCP settings (.cursor/mcp.json):

{
  "mcpServers": {
    "parallect": {
      "command": "npx",
      "args": ["-y", "@parallect/mcp-server"],
      "env": {
        "PARALLECT_API_KEY": "par_live_your_key_here"
      }
    }
  }
}

Claude Desktop

Add to your Claude Desktop config (~/Library/Application Support/Claude/claude_desktop_config.json):

{
  "mcpServers": {
    "parallect": {
      "command": "npx",
      "args": ["-y", "@parallect/mcp-server"],
      "env": {
        "PARALLECT_API_KEY": "par_live_your_key_here"
      }
    }
  }
}

Claude Code

claude mcp add parallect -- npx -y @parallect/mcp-server

Then set PARALLECT_API_KEY in your environment.

Transport Modes

The server supports two transport modes: stdio for local MCP clients and HTTP for hosted/remote deployments.

Stdio (Local)

Used by Cursor, Claude Desktop, Claude Code, and other local MCP clients. The API key is set via environment variable and used for all requests.

Hosted HTTP

Runs as a standalone HTTP server. Each client authenticates with their own Authorization: Bearer <token> header, which is passed through to the Parallect API.

# Start the HTTP server
PORT=8080 PARALLECT_API_URL=https://parallect.ai npm run start:http

Clients send MCP JSON-RPC messages via POST /mcp with their API key as a Bearer token:

curl -X POST http://localhost:8080/mcp \
  -H "Authorization: Bearer par_live_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","method":"initialize","params":{...},"id":1}'
Variable Required Default Description
PORT No 8080 HTTP server listen port
PARALLECT_API_URL No https://parallect.ai API base URL

No PARALLECT_API_KEY env var is needed in HTTP mode — each request carries its own token.

Configuration (stdio mode)

Variable Required Default Description
PARALLECT_API_KEY Yes Your Parallect API key (par_live_...)
PARALLECT_API_URL No https://parallect.ai API base URL (for self-hosted or dev)

Get your API key at parallect.ai/settings.

Available Tools

Tool Description
research Submit a deep research query across multiple AI providers
research_status Check progress of a running research job
get_results Retrieve the synthesized report, citations, and claims
follow_up Pursue follow-on research in the same thread
list_threads List your recent research threads
get_thread Get a thread with all messages, jobs, and results
balance Check your credit balance and recent transactions
usage View spend analytics by provider and time period
list_providers See available research providers and budget tiers
search_claims Search and filter extracted claims
get_claim_evidence Get the full evidence chain for a specific claim

Development

npm install
npm run build

# stdio mode
npm start

# HTTP mode
npm run start:http

Or for development with auto-reload:

# stdio
npm run dev

# HTTP
npm run dev:http

License

MIT

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选