master-fetch
MCP server for web fetching with Cloudflare bypass, Trafilatura extraction, and smart routing. Free, self-hosted, no API keys.
Hound is a comprehensive MCP server that gives AI agents complete web research capabilities through a single local process. It combines anti-bot web fetching with Cloudflare bypass, best-first crawling with sitemap support, keyless local search across 10 backends with neural reranking, and PDF extraction with OCR for scanned documents. The server runs entirely on your machine with zero API keys, no accounts, and no per-request costs, making it ideal for agents that need reliable web access without external dependencies or billing.
Key Features
Use Cases
- 01Research agents gathering information from multiple sources with automatic search, fetch, and crawl coordination
- 02Documentation crawlers extracting structured content from technical docs with sitemap discovery
- 03Academic paper analysis with automatic OCR recovery of CID-corrupted PDFs and table extraction
- 04Reddit and forum research with optimized extraction and promoted content filtering
- 05Multi-source fact-checking using cross-backend consensus ranking and relevance scoring
- 06Privacy-focused web research with local keyless search and no third-party data routing
Related MCP Servers
View moremarkitdown
Python tool for converting files and office documents to Markdown.
firecrawl
The context API to search, scrape, and interact with the web at scale. 🔥
prompts.chat
f.k.a. Awesome ChatGPT Prompts. Share, discover, and collect prompts from the community. Free and open source — self-host for your organization with complete privacy.
langflow
Langflow is a powerful tool for building and deploying AI-powered agents and workflows.
master-fetch — FAQ
What is Hound and what does it do?+
Hound is an MCP server that provides AI agents with web fetching, crawling, search, PDF extraction, and screenshot capabilities. It runs entirely on your machine with zero API keys or accounts required, combining 10 keyless search backends, Cloudflare bypass, and automatic OCR for scanned PDFs.
How do I install and connect Hound to my AI agent?+
Install with 'pip install hound-mcp[all]' followed by 'playwright install chromium'. Then add a new MCP server entry to your client's config file with command 'hound' and no arguments. After restarting your client, the 6 Hound tools will be available.
Which AI clients work with Hound?+
Hound works with any MCP-compatible client including Claude Code, Claude Desktop, Cursor, OpenCode, Hermes, and Pi. It also supports HTTP mode for Open WebUI v0.6.31+ and has a native Pi extension.
Do I need API keys or a proxy to use Hound?+
No API keys are required for core functionality—search is keyless and local across 10 backends. Optionally, you can add your own API keys (BYOK) for Serper, Tavily, Exa, Firecrawl, or TinyFish, or configure proxies for sustained heavy use with 'hound proxy add'.
Is Hound free to use?+
Yes, Hound is completely free with an MIT license. There are no accounts, no per-request billing, and no data routed to third-party services. All search and extraction runs locally on your machine.
What are the system requirements and limitations?+
Hound requires Python and works on all platforms. The full install with browser support needs Playwright's Chromium. It's designed for agentic research, not bulk scraping at scale, and cannot bypass advanced anti-bot systems like interactive DataDome or Akamai. Sites requiring login are out of scope.
How do I install master-fetch?+
Open the source repository on GitHub and follow its README. master-fetch is a mcp server — MCP Agents Market links you directly to the official repo.
Is master-fetch free?+
master-fetch is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.