Local-first MCP server for web search, crawl, extract, cache, and research without API keys, with optional LLM integration.
Key Features: Multi-engine web search with local ML reranking JavaScript rendering for dynamic web pages Persistent local SQLite cache for all data Self-hosted unified web layer for AI agents offering search, stealth browsing, auth, and platform actions with 5,000 free Gemini searches/month.
Key Features: Search via Gemini Grounded and fallback engines with 5,000 free monthly queries Stealth browsing with Playwright, Camoufox anti-fingerprint, and auto-escalation Multi-page crawling and structured data extraction Web scraping with stealth HTTP, real browsers, and Cloudflare bypass. CSS selectors supported.
Fetch and convert webpages to clean Markdown with link preservation, caching, and polite crawling.
Key Features: Content extraction using Mozilla Readability HTML to Markdown conversion with Turndown and GFM support Smart caching with SHA-256 hashed URLs MCP server for Dumpling AI APIs offering data extraction, web scraping, document conversion, AI agents, and code execution.
Key Features: Access to Dumpling AI data APIs for search, maps, news, and reviews Web scraping with rendering, crawling, screenshots, and extraction Document conversion including OCR and multimedia processing Search the web, fetch pages as clean Markdown, and synthesize multi-source research through MCP.
Key Features: Web search with Brave Search or DuckDuckGo fallback Page extraction in article, full-body, or links-only modes Multi-step research with relevance scoring Node MCP server providing Kagi search and summarization via user session token without official API access.
Key Features: Search Kagi with multiple concurrent queries Summarize URLs with paragraph or bullet point output Supports webpages, videos, and audio content summarization MCP server providing Google News search with multi-language support, topic categorization, and SerpAPI integration.
Key Features: Query-based and topic-based news search Automatic topic categorization of results Multi-language and multi-region support Fetch and search Unsplash photos with full attribution and download tracking for LLM content generation.
Key Features: Photo search with filters (color, orientation, content safety) Random photo retrieval with optional keyword filtering Download tracking endpoint for compliance Open-source MCP server for bounded, multi-engine web search and URL extraction without API keys.
Key Features: Eight free search engines with waterfall fallback Bounded search budgets and quality gates Multi-source evidence with provider provenance Fetch web page content through FourA using automatic selection, direct HTTP, proxy rotation, or browser rendering.
Key Features: Automatically selects a fetch method Supports HTTP, proxy rotation, and browser rendering Returns typed structured content Headless browser MCP server capturing tiled 1072x1072 screenshots and element snapshots optimized for AI vision workflows.
Key Features: Automatic tiling of full pages into 1072x1072 pixel chunks Supports full page and element-specific screenshot capture Configurable viewport size and wait strategies for dynamic content 
Unsplash Mcp