ollama-mcp

Ollama MCP Server connects your local language models to Claude Desktop, Cursor, Cline, and other MCP-compatible AI assistants through the Model Context Protocol. It exposes 14 tools covering model management, text generation, chat interactions with tool/function support, embeddings, and web operations. The server supports both local Ollama instances and Ollama Cloud, enabling hybrid deployment where you run local models privately while accessing cloud-only features like web search and content fetching. With zero dependencies, TypeScript type safety, 96% test coverage, and automatic tool discovery, it provides a secure, performant bridge between your AI agents and Ollama models without leaving your infrastructure.

["Install and configure the ollama-mcp server on your local machine. Follow the [official Ollama MCP documentation](https://github.com/ollama/ollama-mcp) to set up the server and ensure it’s running.","In your AI assistant (e.g., Claude, Cursor), enable the ollama-mcp server as a tool. For Claude, add it to your `claude_desktop_config.json` under the `mcpServers` section.","Use the prompt template to craft your request. Replace [TASK] with your specific goal (e.g., 'analyze,' 'summarize,' or 'generate'). Include [PLACEHOLDERS] like file paths, models, or context.","Execute the command or query in your AI assistant. Monitor the output for accuracy and refine your prompt if needed. For complex tasks, break them into smaller steps.","For iterative workflows, save frequently used prompts as templates in your AI assistant’s custom instructions or a notes file to streamline future use."]