← Back to directory
COMPARE UP TO 4 SERVERS

Compare MCP servers

Compare scores, permissions, risks, and fit in one decision-focused table.

DimensionLocal Deep Research (LDR) MCP ServerLocal-first deep research assistant exposing multi-engine search and cited reports to Claude over MCP.Firecrawl MCP ServerFirecrawl's official MCP server for web search, scraping, and structured extraction for AI agentsExa MCP ServerExa's official MCP server for web search and crawling built for AI applications
FMRS70 / 100 · B75 / 100 · B73 / 100 · B
Reliability9 / 2013 / 2012 / 20
Security and permissions14 / 2012 / 2013 / 20
Maintenance18 / 2016 / 2016 / 20
Documentation17 / 2017 / 2016 / 20
Setup experience12 / 2017 / 2016 / 20
Best for
  • Privacy-conscious researchers and developers who want to run LLMs and search fully locally
  • Users already running local models via Ollama, LM Studio, or llama.cpp
  • People who want to give Claude multi-engine academic search and structured report generation
  • Local, STDIO-based integration into Claude Desktop or Claude Code
  • Scenarios needing clean, structured web context fed to an AI assistant
  • Users already on the Firecrawl platform who want to call its scraping capability directly via MCP
  • Q&A and research scenarios that need real-time, clean web content
  • Users with an existing Exa account who want a zero-deployment remote option
Not for
  • Multi-user or network-exposed deployments — this MCP server has no built-in authentication or rate limiting and is documented as local-use only
  • Users wanting a zero-config tool — it requires configuring an LLM provider and a search engine, and sometimes deploying SearXNG
  • Teams needing real-time multi-user collaboration features
  • Simple static-page scraping where you don't want to depend on a third-party API and incur call costs
  • Sites that explicitly disallow automated access (robots.txt)
  • Scenarios requiring a strict allowlist of search sources that can't use a third-party search service
  • Budget-sensitive, high-volume search scenarios (remote/API calls may incur cost)
Required permissions
  • Access to a configured LLM (local endpoint or cloud API key)
  • Outbound network access to call configured search engines (arXiv, PubMed, SearXNG, Brave, etc.), unless restricted to local documents only
  • Local filesystem read access for analyze_documents to work with private documents
  • Read/write access to the local encrypted (SQLCipher) database
  • Requires a Firecrawl API key to call; cost and quota are governed by the Firecrawl account
  • firecrawl_agent/firecrawl_interact perform automated browser interaction, which may trigger login or form-submission flows on the target site
  • Requires an Exa API key for full functionality; OAuth login is optional in remote mode
  • web_fetch_exa retrieves content from a specified URL; agent_run performs multi-step autonomous actions and needs the user to have clear expectations of Agent behavior
Risks and side effects
  • The docs explicitly warn this MCP server is designed for local use only via STDIO transport, has no built-in authentication or rate limiting, and should not be exposed over a network without additional security controls
  • Credentials (e.g. API keys) are held in process memory during active sessions, a routine runtime risk the project mitigates with session-scoped credential lifetimes and core dump exclusion
  • Research queries are sent to whichever search engines are configured, and cloud LLM providers (OpenAI, Anthropic, Google, etc.) receive query/document content when selected
  • Docker's --network host mode can silently fail to expose ports or misroute localhost on Windows/WSL2/Mac
  • Bulk crawl/map tools can generate significant request volume against a target site — respect the site's rate limits and terms of service
  • firecrawl_agent's interactive action chain is longer — define task boundaries clearly before running it to avoid accidentally triggering actions on a sensitive site
  • The API key should be stored carefully — avoid leaving it in plaintext in client config files on untrusted machines
  • agent_run is a multi-step autonomous tool with a longer action chain — try it on low-risk tasks first
Supported clientsClaude Desktop, Claude CodeClaude Desktop, VS Code, Cursor, Windsurf, Zed, AmpClaude Desktop, Claude Code, VS Code, Cursor, Windsurf, Zed, Amp, Kiro, LM Studio, Antigravity
Tools8114