MCP Find
Browse Servers
Blog
Docs
Submit Server
Back to Directory
/
AI & ML
deadends.dev
Structured failure knowledge for AI agents — dead ends, workarounds, error chains
AI & ML
v0.3.1
View on GitHub
Learn More
How to Use MCP Servers With Ollama and Local LLMs
Ollama does not natively speak MCP, but bridge tools like MCPHost let local models call MCP tools over stdio. MCPFind indexes 832 servers in the ai-ml category. This guide covers which models support tool calling, how to wire MCPHost to any local model, and which server categories are worth running offline.
Read article
MCP vs LangChain Tools: Which Should You Use?
LangChain tools and MCP servers both give AI models access to external functions, but they work at different layers. This guide compares them on architecture, cross-client reuse, and long-term maintenance using adoption data from the MCPFind directory's 6,105 indexed servers.
Read article
MCP vs Function Calling: Which Should You Build With?
MCP and function calling solve different problems in AI agent architecture. Function calling is tight, in-process, and fast. MCP is modular, reusable across models, and designed for production tooling. We analyzed both patterns against real servers in the MCPFind directory to map out when each one wins.
Read article